Nozzle spray angle...

[jondee86]

I am looking at possible injector locations and would like to set the nozzles
so that there is no projection into the airflow, and have the tip of the
nozzle recessed 2-3 mm as per this drawing...

http://www.aquamist.co.uk/HFS4/wiring/Ford%20Focus/285+357N.jpg

So if someone can advise me the injector spray cone included angle, I can
look at how this setup can be achieved on 1.6-2.0 mm wall alloy tube.

Also, I have seen some data that suggests recessing the tip promotes
better vapourisation, and I would appreciate any input indicating that this
is a worthwhile exercise.

Cheers... jondee86

PS: I do recall seeing a post that mentioned the cone angle was 80 deg,
but I have been unable to locate that again.

[jondee86]

Ha... found a post that mentioned a spray cone angle of 80 deg so I guess
that answers part of my question. And while I am on the subject of nozzles
there is also the question of preferred placement for best performance.

I am putting a Sprintex S5-210 onto a 16V 4AGE engine and this pic shows
the intake and discharge ducting I will be using...

http://i984.photobucket.com/albums/ae328/jondee1/Sprintex%20ducting%2001_zpskp3yuuzu.jpg

The Intake is on the right, and comprises a cast manifold with some 2-1/2"
alloy tubing. The throttle attaches to the open end of the tube. Proposal
is for one small nozzle pre-charger.

The discharge is on the left and comprises a cast manifold plus 2-1/2" alloy
tubing with three 90 deg bends. The small transition will connect the last
bend to a stock 4AGZE inlet manifold. Proposal is for one small(ish) nozzle
post-charger.

The general arrangement of the assembled pieces is as per this pic...

http://i984.photobucket.com/albums/ae328/jondee1/Sprintex%20ducting%2002_zpsv495wfzm.jpg

To date my searching tends to indicate that it is OK to locate the pre-
charger nozzle close to the charger inlet, and that the post-charger nozzle
should be as far from the inlet manifold as possible. If anyone has any
suggestions based on their personal experience of what works best, please
post them up here :)

Cheers... jondee86

[parmas]

Hi Jondee,

Your setup looks sweet although for my tastes there is some extra bending there which you could experience air turbulence. If the charger was on the side of the engine going into a 45degree angle to the manifold would be ideal. Although it does not mean it will not work well eventually :) .

Why would you make such an impact on nozzle angle ? Some setups have nozzles pointed directly to the flow coming and work well. I have mine in a 45Degree bend.

Why hiding it between too walls ? This way injection could have more probability wetting the walls instead of mixing into the air causing liquid water to form into the intake, which is bad...

I attached where would I put the nozzle location PRE/Post Charger

[jondee86]

Thanks for the suggestions parmas

The layout of the ducting was dictated to a large extent by the lack of
space on the intake side of the engine, and by the fact that I did not want
to get involved with building a custom intake manifold at this stage. It was
suggested that I could eliminate all but one of the 90 deg bends on the
discharge side if I entered the GZE manifold from below. While that may
have been possible, I felt that it would starve cylinders 1 and 2.

So I went with a layout that is as close as I could get to the factory angle
of entry into the top of the manifold. I have faith that the Toyota factory
engineers did some testing on the non-symetrical runner layout to make sure
that it worked reasonably well. I have also increased the throttle size to
60 dia and all tubes to 2-1/2" to try and keep flow losses down. This is my
first venture into supercharging, so I am more interested in creating an
engine that makes driving a pleasure, rather than ultimate power.

Mounting the pre-charger nozzle in the bend seems reasonable. I think that
I would mount it square to the surface of the tube to try and project the
mist as far out into the airstream as possible (penetrate the fast moving
layer of air on the outside of the bend).

The post charger nozzle where you suggest, or perhaps in the straight
tube below the small angle bend to place it out of the way of mechanical
damge. Again installed square to the tube surface.

My interest in recessing the tip of the nozzle comes from reading this page..
http://www.alcoholinjectionsystems.com/article_info.php?articles_id=100

http://www.alcoholinjectionsystems.com/images/water_injection_nozzle_54.jpg

Allowing the cone to develop before exposing it to the "shearing" effect of
the airstream seems logical to me. It will still get sheared, but the droplets
will be finer and should disperse more effectively. Sound reasonable ?

Cheers... jondee86

[Richard L]

Looking at the angle of those nozzle, it could do with every little help to improve the atomisation. Take a look of the video below:

https://www.youtube.com/watch?feature=endscreen&NR=1&v=xvSNmn6A64M

[jondee86]

Oh, I don't know what type of nozzles they are... but obviously not yours :)
The image was photoshopped simply to show how recessing the nozzle tip
would allow the cone to expand a little more before it was "folded over" by
airflow within the tube. I assume that even your nozzles have the water
leaving the pinhole as a solid stream which very quickly expands and breaks
up into small droplets.

My interpretation of this information is that the water leaving an exposed
nozzle would be more likely to be folded back against the wall of the tube
than the same nozzle recessed a few mm below the tube surface. However,
I'm sure that you have more than likely tested this theory in the lab. And
if experimentation shows that there is no benefit from recessing the nozzle
tip, then I'll just use the standard adapter and try and get the end of the
nozzle flush with the inner surface of the tube.

Cheers... jondee86

[parmas]

That's why I like to put an injector in the middle of an elbow of 45degree since putting it just into a straight pipe could lead to wetting walls/inefficient dissipation.

Although I would be interested if someone prove scientifically what works best

[jondee86]

Personally, I am inclined to think that placing the nozzle in the tube just
before the inside of a bend would result in the best mixing.

https://encrypted-tbn2.gstatic.com/images?q=tbn:ANd9GcQ7Ty4AKSVkfr3JGguoKNY 9kFcD6mjTpJivD6zbBHxr5DBZ30tDJg

Using this simulation as a reference, and assuming that the spray is swept
away and remains quite close to the nozzle side wall, the spray would then
be turned around the inside radius of the bend where flow separation and
turbulance occurs. This should promote better dispersion and mixing of the
air and water.

However, this remains just a theory without any valid indication of exactly
how the spray penetrates the airstream once it leaves the nozzle.

Cheers... jondee86

[jondee86]

Been doing some searching but so far have not been able to find anything
of much use. However I did find this...

http://i984.photobucket.com/albums/ae328/jondee1/Droplet%20size%2003_zps9zruwy3u.jpg

It shows, as might be expected, that larger droplets penetrate further into
the airstream. Smaller droplets are the first to be swept away. This leads me
to believe that placing the nozzle either on the inside of a bend or immediately
after a bend, would be advantageous. The larger droplets would then be
projected towards the higher velocity air on the outside of the bend, which
would help break them up and prevent/reduce any impingement on the
farside wall.

Cheers... jondee86

[rotrex]

Flow velocity is the fastest on the inside of a bend.
Droplets do not like to take bends. They will hit the wall and wet it.
http://www.appliedmc.com/images/e-zine/flow-through-elbow.jpg
This is the reason fuel injectors are spraying straight into the inlet port.
They are not mounted before any bends. In the old days of throttle body injection the manifolds were first designed for wet flow and they were heavily heated to let that wet fuel film on the intake plenum wall evaporate.

Water evaporates about 10x slower than methanol and much much slower than gasoline.
If you want water meth as droplets in the cylinder and not as a stream along the wall you have two choices.
Either have the nozzles in the straightest section before the inlet ports of the engine or mount the nozzle as far away from the engine as possible. In the latter case there is more time to evaporate, even if streaming along the wall. In the first case, the highest percentage of droplets will make it into the cylinder.

Calculate the g-forces in a bend of your radius at WOT and max flow. You will find numbers of 1000G and more. These are typical numbers found in vortex water separators to remove water droplets from natural gas.

In the end all loss of fluid can be compensated by injecting more or use of a higher percentage of methanol. It evaporates on the wall as well as in the air.

IMHO inject as suggest by parmas straight into the SC and then in the straight section leading into the intake plenum. Your center cylinders will receive more spray than the outer ones as the drops have to take more bends. Little you can do about it except going for a direct port set-up.

[jondee86]

Thanks for the input rotrex :)

If the water is more effective in the cylinders as droplets rather than
vapour, then it makes sense to move the post-charger nozzle closer to
the inlet manifold. I'm not sure which position you are recommending, so
I have marked them here...

http://i984.photobucket.com/albums/ae328/jondee1/Sprintex%20ducting%2003_zpsrzay9sk4.jpg

On top of the last bend would give a straight shot into the manifold, and
probably separate a fair amount of water out on the manifold floor. However
this will find its way into the cylinders no problem as the engine is installed
with a slight slope.

Will it make any significant difference if the nozzle is mounted radially to
the bend, or axially to the following tube ? Same situation applies to the
pre-charger nozzle.

Cheers... jondee86

[rotrex]

I'd take the right position, maybe a bit further left, say 2cm, so the spray mostly points into the pipe. You might even point the nozzle straight down into the pipe towards the manifold.

The left nozzle position would likely lead to a large percentage of the droplets hitting the outer walls of the pipes creating more of a wall bound fluid stream than mist.

Water vapor does little for you power wise, it actually costs power. You need it mostly as fine fluid droplets in the cylinder.
For methanol this is less of an issue. Most will evaporate even just hitting the hot inlet valve. There has been research done on comparing ethanol port injection vs direct (in cylinder) injection regarding charge cooling effect.
http://www.aquamist.co.uk/forum2/vbulletin/showthread.php?t=2872&referrerid=6624
They found that much of the ethanol seems to evaporate hitting either the inlet valve or the cylinder walls (cylinder, head and piston crown) before the intake valve closes. It rather cools the metal than the inlet charge and does little to raise the knock limit by means of in cylinder charge cooling. It still raises knock limit due to its chemical properties aka high octane number.


That can be a good thing, too, say for a heat sensitive motor lacking piston oil jet cooling or a rotary engine.

One more suggestion. Mount nozzle threats for both positions and compare what gains you the most power after ignition tuning. You will find that the ignition advance requeirements will be rather different for the two positions. The position that needs the most ignition advance to restore full power got the most water mist into the cylinder. With the engine being out of the car this is very easy to implement.

The more methanol your mix contains, the less the nozzle position matters.

[parmas]

Nice explained Rotrex. I like your thinking style makes me wonder of an updated revision on mine for my next high compression build...

[jondee86]

OK... I think I'll go with the in-line (back of the bend) location. Should be
easy enough if I can use these nozzles...

https://lh3.googleusercontent.com/Fdlml4T_-Rh4qmoMjk6AU2BUWKZ-OFMShrS5cStkAAQ=w417-h741-no

As suggested, I will position the nozzles a little bit towards the incoming
airflow to try and keep as much water off the walls as possible. Having the
tip of the nozzle projecting a few mm should help a little as well.

@ Richard L
I don't see the extended nozzles anywhere on the Aquamist Direct site. Can
you confirm that they are selected the same as standard nozzles ?

Cheers... jondee86

[rotrex]

point it towards the opening of the intake plenum, so straight down

[jondee86]

I'm thinking something along these lines would work ?

http://i984.photobucket.com/albums/ae328/jondee1/Nozzle%20location%2002_zpspbmssfbb.jpg

Should be easy enough to make a couple of aluminum bungs and have
them welded to the tube. I don't have a lot of clearance on the intake
ducting, so I would have to check that there is enough room for the longer
mounting detail. Right angle fittings should work.

EDIT: Due to the amount of messing around involved in mounting the
nozzles in-line, I am thinking that I will just put them at 45 deg as shown
in the revised sketch. Since 99% of people mount them straight in the
side of a duct and it works, mounting in the corner should also work.

Also means I can use standard nozzles with a button welded on the tube
and tapped M8.

Cheers... jondee86

[jondee86]

@rotrex Actually, looking back at the advice you gave in this thread...
http://www.aquamist.co.uk/forum2/vbulletin/showthread.php?p=21006#post21006

I am starting to think that port injection may be the easiest/best solution
to ensuring even distribution of water. This picture shows that there is more
room under the fuel rail than I first thought, and it should be possible to tap
a nozzle into the top of each runner close to the injector.

http://i984.photobucket.com/albums/ae328/jondee1/Nozzle%20location%2003_zpsflhje0ul.jpg

This location will avoid water pooling on the manifold floor, and bypasses the
question of how best to mount a nozzle in the ducting. My research to date
has not found any urgent reason to spray pre-charger, other than when
using the car for sustained high speed operation (racing). So I will probably
put that on hold for now. It would be a simple job to add a 5th nozzle at a
later date.

I have a couple of questions regarding the best nozzle and fitting setup for
port injection...
1. Do I need checkvalve nozzles ?
2. If so, what is the lowest height option for a right angle connection ?

Cheers... jondee86

[Richard L]

@ Richard L
I don't see the extended nozzles anywhere on the Aquamist Direct site. Can
you confirm that they are selected the same as standard nozzles ?

Cheers... jondee86

The extended nozzle is only available in the "internal checkvalve (http://www.aquamist-direct.com/806-501a-to-806-528c-checkvalve-jet-for-4mm-hose-qty-discount-designed-specially-for-direct-port-applications/)" version.

.

[parmas]

@rotrex Actually, looking back at the advice you gave in this thread...
http://www.aquamist.co.uk/forum2/vbulletin/showthread.php?p=21006#post21006

I am starting to think that port injection may be the easiest/best solution
to ensuring even distribution of water. This picture shows that there is more
room under the fuel rail than I first thought, and it should be possible to tap
a nozzle into the top of each runner close to the injector.

http://i984.photobucket.com/albums/ae328/jondee1/Nozzle%20location%2003_zpsflhje0ul.jpg

This location will avoid water pooling on the manifold floor, and bypasses the
question of how best to mount a nozzle in the ducting. My research to date
has not found any urgent reason to spray pre-charger, other than when
using the car for sustained high speed operation (racing). So I will probably
put that on hold for now. It would be a simple job to add a 5th nozzle at a
later date.

The Aim of injecting Pre-Charger is to increase the efficiency of the charger if by any means you are already working out of it's efficiency range. Also for setups without intercooler (like me) that don't have long piping giving time to evaporate and cool the air mixture. Tried and tested and works without problems.

I have a couple of questions regarding the best nozzle and fitting setup for
port injection...
1. Do I need checkvalve nozzles ? RECOMENDED
2. If so, what is the lowest height option for a right angle connection ?

Cheers... jondee86

Having the setup port injected will not affect the air mixture thus no cooling is being made. I would still make at least one nozzle post charger to gather that extra hp from a denser mixture.

[jondee86]

Hahahahahah... just had a look at what a port injection system costs :eek:
So an end to thinking out loud. It is always possible to upgrade, so for now
I will go with a basic two nozzle system... one pre-charger and one post-
charger. Standard nozzles installed at 45 deg in the bends as per the above
sketch.

From what I understand, placing the post charger nozzle further from the
inlet manifold would help the cooling effect, but increase the chances of
water being centrifuged out of the airstream. Placing it in the last bend will
reduce the time for cooling, but allow more water into the cylinders in
droplet form... so perhaps the better option ?

The pre-charger nozzle appears to be less critical as the charger "mixes"
the air and will not be sensitive to droplet distribution at the inlet.

The only question remaining is whether it is necessary to use one (or two)
checkvalve nozzles ? There will be some pressure differential between the
nozzles, so draining could occur during out of boost driving. However, there
will only be about 0.5 metres of 4 mm tube between the nozzles, and this
should not create any significant delay between activation and spray ?

EDIT: Oh,and one last question... what would be the recommended water
split be between the pre and post charger nozzles ? 25/75 ?

Appreciate the input as I like to examine the alternatives :)

Cheers... jondee86

[parmas]

Hahahahahah... just had a look at what a port injection system costs :eek:
So an end to thinking out loud. It is always possible to upgrade, so for now
I will go with a basic two nozzle system... one pre-charger and one post-
charger. Standard nozzles installed at 45 deg in the bends as per the above
sketch.

From what I understand, placing the post charger nozzle further from the
inlet manifold would help the cooling effect, but increase the chances of
water being centrifuged out of the airstream. Placing it in the last bend will
reduce the time for cooling, but allow more water into the cylinders in
droplet form... so perhaps the better option ?

The more the consistency of the injection the less variables you will get. Droplets are only good for cooling the charge while injection is stopped or throttle is closed. The cooling effect is so drastically seen since engine is at no load/off boost (decelaration) and heat is not generated while water is still evaporating into the piping. How much? From 40-15DegC in seconds easily.

As in regards to positioning of the nozzles, I already gave my one cent.

The pre-charger nozzle appears to be less critical as the charger "mixes"
the air and will not be sensitive to droplet distribution at the inlet.

The only question remaining is whether it is necessary to use one (or two)
checkvalve nozzles ? There will be some pressure differential between the
nozzles, so draining could occur during out of boost driving. However, there
will only be about 0.5 metres of 4 mm tube between the nozzles, and this
should not create any significant delay between activation and spray ?

Again, as already explained before if your charger matches your engine and working into it's efficiency eliminate the pre-charger nozzle. Regarding nozzles, I like built in checkvalve nozzles but that is your choice.

EDIT: Oh,and one last question... what would be the recommended water
split be between the pre and post charger nozzles ? 25/75 ?

Trials is your best answer. I would begin with water only to eliminate excess fueling and AFR changes. After that trial, report back to us and we will try to help you in the best way possible

Appreciate the input as I like to examine the alternatives :)

Cheers... jondee86

Jondee, refrain from asking more questions, you have a good thread to refer on. Read it out slowly and decide from where to begin.

Good Luck!

[rotrex]

Port injection costs three extra jets, that is all.
You can get 4 jets with integrated check valves form aquamist and combine them with any feed of pressurized mix you like.
To test the post supercharger jet, just get a extra jet for this and as parmas said, try it out.

My system is a former Auqamist 2c. After I had pump issues, I installed a Flotec 160PSI pump and slightly modified the aquamist loom to maintain functionality, .ie. Blue LED and system off if tank is empty and to maintain the pump runs indicator.
I kept the constant presure supply and the Aquamist fast acting valve controlled by the ECU. The valve does not care what jets I mount. :-) As long as the pump is up to it, it works, The pump itself runs to its preset pressure and stops. Should the railö pressure drop due to injection, it automatically starts again. It also triggers a LED in my cockpit. The faster it flickers, the more I inject. :-)

With water you will lose power unless you advance the ignition timing. The more you drop power, the more water enters the cylinder in the form of mist.

Regarding jets, I'd purchase a selection. For the port injection, I am on 4x 0.4mm C aquamist jets and have the feeling I could even go to 4x 0.5mm. Literally, the more I inject, the more power I can extract after a ignition adjustment.


If your ECU has a adaptive fueling option, just give it your desired target AFR and let it sort the fueling with the extra alcohol injected.

If you start with water, you'll find the engine will sound soft. This is a indication of the slower and delayed burn.

In terms of space, you need some 3cm above the plenum for the jet, a piece of 4mm tubing and a 90 deg bend.

Other manufacturers may have flatter jets or some that are mounted to a 90 deg bend.
I have good experience with the Aquamist jets. You can maintain them as they are screwed together from two pieces, never had once blocked up, contain a filter fritt. Always use a pre filter. No component of any WI system likes any dirt or mineral deposits. Always use deionized water. If it's marked for drinking, it is no good as RO purified water for drinking purposes has minerals added for tast and health.

[jondee86]

Thank you gentlemen :) There is a lot to learn and some experimentation
will be required to gain more understanding. I shall start with a single nozzle
and add a second once I have the system installed and working for long
enough to get the basic tuning completed.

I shall also be using different pulleys so that I can start with 10psi boost,
move to 15psi, and then finally to 20psi if the results indicate that 20psi can
be reached with my setup. Adding port injection will be something to consider
when I change to the 15psi pulley.

It will be several months before the conversion from NA to supercharged is
completed, but you can be sure I will report back once I have the car tuned
and the WI system working.

Cheers... jondee86

[rotrex]

15 and especially 20 PSI are some serious boost numbers to push through a roots style blower without charge cooling and on petrol.
As Parmas and others did you will indeed need a large amount of ADI fluid to keeps this from having knock issues.
I also suggest in this case to consider pre compressor injection as suggested ans successfully employed by Parmas. Thing will get really hot.
Have you checked your compressor map regarding the expected flow and boost numbers?
Are you still reasonably efficient?

[jondee86]

Ahh... good questions, and I don't have any good answers :o

When I began this project I was only intending to target 15psi, but the
only pulley I could get at the time was a small one that was 20psi capable.
I did look at A2W intercoolers, but decided to see what was possible with
WI because it was easier to implement. These are the maps I was given...

Isentropic Efficiency
http://i984.photobucket.com/albums/ae328/jondee1/S5-210%201.4PR%20isen%20map%2002s_zps4rxgwf e2.jpg

Volumetric Efficiency
http://i984.photobucket.com/albums/ae328/jondee1/S5-210%201.4PR%20VE%20map%2002s_zpsv3yx1yeb .jpg

I don't really fully understand how to interpret them. Hopefully, being a
modern twinscrew design, WI will be able to control the heat generation
up to some reasonable pressure. I am not obsessed with reaching 20psi
and will be quite happy if I can run thru the gears at 15psi now and then :)
The supercharger I am using has an internal pressure ratio of 1.4 and my
engine has a compression ratio of 8.9:1 if that helps ?

Cheers... jondee86

[rotrex]

What's the HP target? This SC seems good for some 300HP, so about 250g/sec of air, for a pressure ratio of 2, so a tad less than 1 bar of boost. This suits to a engine with NA power of about 150HP.

The first diagram shows the efficiency over airflow and pressure ratio. You need about 100g of air per second for 120HP. The lower the efficiency the hotter the air gets out of the SC. The second map shows what percentage of the theoretical volume of the SC is expelled per revolution. Say it pumps 1liter per revolution and vol efficiency is 89%, you only get 890ml of air out per revolution. Pressure ratio is the ratio of output pressure and input pressure, both on the absolute scale. So athmospheric, 1bar absolute, in and 2 bar absolute or 1bar of boost out is 2/1=2. There always intake losses, so your real PR is always a tad lower (10-15%) than the ideal number.

[parmas]

Ahh... good questions, and I don't have any good answers :o

When I began this project I was only intending to target 15psi, but the
only pulley I could get at the time was a small one that was 20psi capable.
I did look at A2W intercoolers, but decided to see what was possible with
WI because it was easier to implement. These are the maps I was given...

Isentropic Efficiency
http://i984.photobucket.com/albums/ae328/jondee1/S5-210%201.4PR%20isen%20map%2002s_zps4rxgwf e2.jpg

Volumetric Efficiency
http://i984.photobucket.com/albums/ae328/jondee1/S5-210%201.4PR%20VE%20map%2002s_zpsv3yx1yeb .jpg

I don't really fully understand how to interpret them. Hopefully, being a
modern twinscrew design, WI will be able to control the heat generation
up to some reasonable pressure. I am not obsessed with reaching 20psi
and will be quite happy if I can run thru the gears at 15psi now and then :)
The supercharger I am using has an internal pressure ratio of 1.4 and my
engine has a compression ratio of 8.9:1 if that helps ?

Cheers... jondee86

I am sorry to tell you this but you better leave that 20psi pulley on the shelf.

I am attaching you the best and max high boost efficiency.

The First line shows where the charger is meant to work best which is 7-8psi @ 90g/sec or 12lbs/min. To make it more simple : With my 1.5L engine and a GT2560 turbo, I reach 12lbs/min @ about 3000rpm.

The Second line shows where the charger peak it's efficiency or in poor words begins to "overheat" the Air drasticaly compared to relatively lower pressures.
The Max working pressure is 15psi @ 170g/sec or 22.5lbs/min. To make it more simple : Compared with mine I reach 22.5lbs/min @ 5500rpm

If I knew your engine specs I could understand where this charger is with your engine. Specs needed:

- Engine size cc
- No of cylinders
- No of valves

To simply understand boost efficiency if you make a chart of boost vs air temperature you will understand without any compressor maps where your charger efficiency is. The chart will show elevated/abnormal air temperatures when out of efficiency.

Example :

Air Temp marking 30DegC at 5psi @ 3000rpm and 31DegC at 8psi @ 4000rpm.

DIFF 3PSI = 1DegC increase

Air Temp marking 34DegC at 8psi @ 3000rpm and 40DegC at 11psi @ 4000rpm.

DIFF 3PSI = 6DegC increase

A super efficient compressor would increase Air Temp equally related to boost differential at same RPM. Sadly no compressor is so super at high boost levels ...

[rotrex]

At a pressure ratio of 2 efficiency mostly stays around 70% even at well over 200g/sec. That is OK.

[jondee86]

I'll have to read that a few more times to see if I can understand !!

My engine is 1600cc 4 cylinder 16 valves twincam Toyota 4AGE 8.9:1 comp.
The Sprintex displaces 0.94 litres/revolution with internal PR of 1.4

I don't have a chart for the Sprintex but here is the chart for an Opcon
twinscrew displacing 1.2 l/rev and internal PR of 1.35. Being of similar design
the performance should be similar ?
http://i984.photobucket.com/albums/ae328/jondee1/diagram_lys1200ax_zps49d06897.jpg

And here are typical performance curves for the Opcon Autorotor...
http://i984.photobucket.com/albums/ae328/jondee1/Autorotor%20Performance%20Curves_zpsp7gt 9r6q.jpg

To begin I will have the 10psi pulley and if it is possible to make 10psi with
WI only, I will be happy with that :) If I need to have more I will look to add
A2W using a kit like this...

http://i984.photobucket.com/albums/ae328/jondee1/null_zpsyjiirdln.jpg

Cheers... jondee86

[parmas]

Our specs are similar my engine is 1500cc 4cyl 16valve twincam with 8.5:1 compression.

Forget the charge cooler... You definitely not needing it till 20psi

[jondee86]

Forget the charge cooler... You definitely not needing it till 20psi
Excellent :D Then I shall proceed as planned. I have decided to install three
injection points, all in bends and arranged to spray axial to the following tube.
One will be in the bend before the charger, one in the first bend after the
charger, and one in the last bend before the engine inlet manifold. This will
give the choice of 1, 2 or 3 nozzles spraying simultaneously.

I will also drill, tap and plug the inlet runners (straight up) to allow for the
installation of port injection when it becomes necessary. I have selected one
each of standard nozzles 0.70, 0.50 and 0.40B to begin, and will order more
once initial tuning with water is under way. I think water only should be
sufficient to reach 10psi ? If not, I will change to 25-35% methanol mix.

Once again thanks to you and rotrex for your guidance :)

Cheers... jondee86

[rotrex]

If you have a wideband Lambda Sensor and a programmable ECU, I'd start with a 50:50 mix right away. It requires less changes to ignition timing and works over a wider range of AFRs.

[jondee86]

If you have a wideband Lambda Sensor and a programmable ECU, I'd start
with a 50:50 mix right away. It requires less changes to ignition timing and
works over a wider range of AFRs.
Yes, I have an LC-1 wideband in the car and it will be hooked up to the new
ECU to allow the auto-tune function to work on the fuel map. While I haven't
spent much time considering how to go about tuning the map above 100kPa
I do recall reading that 50/50 v/v was the easy way to mix methanol and
distilled water (deionised not readily available here).

I also recall that tuning with a 50/50 mix comes close to allowing the engine
to run a gasoline only ignition map ? Hopefully, that means I can use timing
from turbo (no WI) ignition maps as a a guide for my initial setup ?

One thing I have been wondering about, is how the WBO2 handles the mix
of methanol and gasoline ? Stoich for meth is 6.4 vs 14.7 for gasoline. So if
the target AFR in boost is 12.5 at a certain load, does the ECU just "ignore"
the presence of water and meth, and simply adjust the gasoline injection
to achieve the target AFR ? Sure makes life easy if it does :)

Cheers... jondee86

[rotrex]

LC-1 is a terrible system. I had two failing in short order. Got a STACK WBO sensor and no more issues ever since.

Start with a turbo map is OK, but retard the ignition before you start in the positive boost "area" even with water methanol.

Forget about AFR for different fuels. The engine likes to run with a certain residual oxygen level in the exhaust gases. This is the same for all fuels. This residual oxygen is converted to a lambda value which is 1.000 for a stoich mix. Then you can convert this lambda to AFR values for different fuels. If you use a AFR signal for feedback to the ECU, just stick the petrol values, I.e. 14.7:1 stoich. Many engines with WI run well between an (petrol indicated) AFR of 11.8 to 12.5. If you go any richter, performance drops a lot. Depending on boost levels, heat, amount of ADI injection etc, they run well even leaner, but things get hotter and knock sensitivity increases. As you do not opt for any charge cooling, I'd stay on the 12-12.5:1 AFR side of things. If you use pure methanol as the sole fuel, you need twice as much compared to petrol, but at best power, your AFR gauge set to petrol will still show some 12:1. So just ignore them. Fuel specific AFRs become important if you gauge your AFR from the fuel flow values. As we have a convenient sensor for this, you do not need to know your actual fuel flow. The ECU also shows it to you as a percent injector duty cycle.

To get around all this better work with lambda values. Second best, just stick to petrol AFRs to judge where you are.
If you add methanol and (petrol indicated) AFR turns richer by 1.5, just reduce fuel until you are back :-)

[jondee86]

Some people like the LC-1 and some people have problems with it. Mine has
been working fine for five years, but I have to admit the car does not get
used a lot. STACK instruments are nice, but they are $$$$$.

I work with AFR instead of Lambda, but the ECU can display either one. At
the moment I am waiting to order a new ECU as the manufacturer is about
ready to release a new model that will have a lot of new features... some
of which will be useful :p I will wait until the end of the month, if the new
model is nor released I will buy one of the older model. It will be good enough
run the supercharged engine and I am already familiar with the software.

Cheers... jondee86

[parmas]

AFR gauges work with voltages. According to the Pre-Set program and voltage the gauge puts a reading.

Example:

Program 1 : Gasoline, 1.25Volts = 12.5AFR

Program 2 : Methanol, 1.25Volts = 5.52AFR

Let's say we use 100% gasoline OFF boost while shifting to 100% methanol ON boost. If program 1 is selected and the engine is tuned to run 5.52AFR on methanol, the gauge should represent a reading of 12.5AFR.

Methanol likes rich mixtures so if using 100% methanol ON boost the lower the AFR value the better power output. Actually seeing 10AFR using program 1 and 100% methanol is considered safe and powerful.

The only drawback is fuel consumption.

[jondee86]

A bit off topic, but I will ask here rather than start a new thread.

I would like to locate an air temp sensor in the intake manifold so that the
ECU can read the temp of air entering the engine. The readily available GM
Delphi open "fast response" sensors seem to be popular for this application.
They are rated for 200kPa and up to 130degC.

http://www.bmotorsports.com/shop/images/iat-sensor2.jpg

What sort of effect does water-meth spray have on the thermistor ?

The GM sensor has a screw-in metal body, but I will be running a 3mm PTFE
thermal insulator gasket on the intake manifold, so manifold heat soak should
not be a significant problem.

Cheers... jondee86

[rotrex]

The mix makes no issues except that you might (I got) get erratic readings and low temp spikes as methanol dribbles evaporate on the sensor bulb.

[parmas]

I have one similar off Haltech. Injected many times few inches in front for years.

Never had issues apart from what rotrex said especially when throttle closed after WOT run

[jondee86]

OK, I'll grab one then. I actually have the original enclosed type sensor
from the 4AGZE (supercharged) manifold. However, I have been told that
the enclosed type of sensor does not react fast enough for non-intercooled
applications. I hope to be able to use the open type sensor reading to trip
an alarm (via the ECU) if the supercharger discharge temperature goes too
high, so fast is better :)

Cheers... jondee86

[rotrex]

You might just pull timing some 20C past the IATs you usually get with a nozzle spraying before the sensor.
There is a IAT ignition correction table in your ECU that can be used as a kind of fail safety.
The same would work for the fuel IAT correction.
This would work on top of closed loop fueling maintaining your AFR if WI fails or injects less than intended.

[jondee86]

Back again ;) Finally got the supercharger installed and the car running but
without the W/I system (I have the HFS3 here waiting to install). I have two
nozzle holders already in the ducting; one close to the S/C discharge and one
in the last elbow before the intake manifold. Here are some pics...

https://s25.postimg.org/anko1afpr/build_thread_61.jpg

https://s25.postimg.org/94koikjrj/build_thread_59.jpg

https://s25.postimg.org/y9bovzj7z/build_thread_57.jpg

At the moment I am driving the car in the <100kPa zone and working on
getting some sense in the fuel map.

Cheers... jondee86

[jondee86]

However, I have found a small matter that needs to be made clear before
the water injection system is installed. I noticed when looking at the log that
I made today, that the ECU was pulling fuel under steady running, so that
the AFR was made leaner than my target AFR :confused:

http://i984.photobucket.com/albums/ae328/jondee1/log%20sample%20001_zpshjkglnvb.jpg

When I investigated the cause, I found that the ECU Manifold Air Temperature
correction table was the reason for the trim. It is the same table that I used
for my N.A. engine. The Aux Temp sensor is located in the S/C discharge
manifold and is only for logging. The MAT sensor is located in the intake
manifold, and is an ECU input. At the time of making this log the intake air to
the engine would have been about 20degC.

So what is the best way to avoid having the ECU pull fuel with a S/C engine ?
Simply set the correction table to zero at all temps and rely on the fuel map
alone to maintain correct AFR's ? Any advice appreciated.

Cheers... jondee86

[rotrex]

Could it be that you exceed the AFR correction range in closed loop operation?
I run my car with +-25% range.
The IAT correction for fuel is similar for NA and FI.

[jondee86]

This is the IAT compensation table I used with N.A. and now with F.I.

http://i984.photobucket.com/albums/ae328/jondee1/IATgraph%2001_zpsluszbivu.jpg

Ambient temps in NZ are typically between 10 and 20degC, maybe up to
25degC on a good day :) So the amount of compensation applied on an N.A.
engine is relatively minor and positive.

With the F.I. engine the IAT is measured after the S/C and under light cruise
running the temp measured at the inlet manifold was 57degC. It is to be
expected that the S/C discharge temp will increase as the manifold pressure
increases. Using the above table this will result in more negative trim being
applied at a time when the ECU will be commanding a richer mixture.

The target AFR table requests more fuel according to boost, and as temp
varies with boost, it could be said that the AFR table calls for more fuel with
increasing temperatures. At the same time, the IAT compensation table is
reducing fuel with increasing temp. Which leads to these two possibilities...

1. Set the IAT compensation to zero and adjust fuel with the target AFR.

2. Allow the IAT table to apply corrections and adjust the target AFR richer
to compensate.

3. Leave the target AFR alone and let the IAT compensation trim the mixture
leaner as the manifold temp rises.

Seems that we are balancing two variables here, and I am wondering if we
can safely reduce this to one variable :)

EDIT: Thinking about this while I was out walking this evening; obviously the
IAT compensation will be useful when misting to lower the charge temp. So
option 3 looks good. And with this option I am going to assume that the AFR
target numbers are nominally based on (say) 20degC ambient temp, so that
fuel can legitimately be trimmed as the charge temp rises. Right/wrong ???

Cheers... jondee86

[rotrex]

AFR is a target the closed loop controller, if active, will try to hit that within it s limits.
IAT compensation compensates for a physical phenomenon. Air density drops about 3% every 10C. Your IAT table compensates for a part of that.

Your IAT is high because the supercharger is hot and heats the air even while cruising.

Fueling is not derived from the AFR table, but from a VE or a pulse duration table.
The AFR table is just there to tell a controller to correct the fueling table in closed loop operation.

with your SC now operative, you first map the fuel map. An indication of this is if your fuel trims are close to 0%, say +-2%. Means your closed loop controller has hardly anything to do.

The ECU also adds fuel proportionally to boost. This works out pretty well if the base feeling is mapped right. also leave this alone.


It seems to me you really need to look into your ECU's operating manual and see how it works. There are several different strategies. Therefore it is hard to give you advice without even knowing what ECU it is, what mode it runs in, if it has a wideband sensor signal fed into the ECU, what the trim value actually means..... (trim can be a correction based on the average % of fuel trim based on a AFR table but could also means corrections based on correction tables, e.g. IAT, barometric pressure, coolant enrichment etc. . So all stuff completely outside of closed loop operation.

[jondee86]

The ECU is an Adaptronic e440d and it does have a lot of tables that adjust
for temperature, but once the engine is up to operating temp all the tables
are zeroed out except for the IAT table. So when manually adjusting the VE
values obtained by the autotune function to get the trim closer to zero, I only
look at the logs where the engine is at operating temp and under steady load.

There is a wideband connected directly to the ECU for monitoring the AFR.
The manual is quite comprehensive but it does not explain some of the finer
points. I am trying to discover if the VE set by the autotune is before the
IAT correction is applied, or after ?

Looking at the logs it seems it is before, as the correction applied by the IAT
table makes the reported AFR leaner than the target AFR. On a N.A. engine
this would be acceptable, but on a F.I. engine I am not so sure.

I'll ask Adaptronic about the actual strategy used.

Cheers... jondee86

PS: I spent some time today researching IAT compensation and have come
to the conclusion that the answer is to set the target AFR a bit richer and let
the compensation table take out fuel according to the manifold temperature.
So long as the adjusted AFR's are in line with the recommended AFR for the
level of boost, I think that this should be fine.

[jondee86]

Small update. Progress has been slow due to numerous small issues that had
to be resolved, and the onset of winter here in NZ. I am investigating why the
s/c is only making 5.16psi @ 7430rpm at the moment, when it should be
making closer to 8.4psi @ 7600rpm.

This engine was sold as N.A. with 10.3:1 compression, and as s/c with 8.9:1
compression. The same camshafts and cam timing were used in both versions.
So I am thinking that maybe there is a little too much overlap for s/c use and
I have retarded the intake cam by 4 deg. This will be tested later in the week
and logged to see if it has any affect on boost.

EDIT: Tested and no apparent change before hitting the limiter.
Bouncing off the limiter did produce another 7kPa and push the IDC up 8% to
80%, but I exclude those values as not representative of a steady pull to
redline.

The other possibility is that the N.A. factory intake/air filter setup has higher
pressure drop now that it is handling greater flow. I am waiting on a new filter
to construct a "free-flowing" intake. This will also be tested once on the car.

And last, when I lifted at 7430rpm the IAT with water only was 54 deg on a
day when the ambient temp would have been no more than 14-15 deg. So if
meth can drop the temp closer to ambient there should be some improvement.

I'd like to get the experimentation over before going to a smaller pulley.

Cheers... jondee86

[rotrex]

For intake restrictions it is best to measure pressure level before the SC intake.
From this value and boost pressure, you can calculate the pressure ratio.
Any cheap vacuum gauge or boost gauge will do. If you see more than 100 mbar of intake restriction, investigate. At a pressure ratio of 2, a 100 mbar intake restriction causes boost to drop by roughly 0.2 bar. 2 = 1.8 bar / 0.9 bar. At 200 mbar, you get 2 = 1.6 / 0.8 = 0.6 bar of boost or 0.4 bar less than without any intake restriction.

Intake restrictions usually show at high air flow, so near max rpm and WOT.


With this and SC rpm you can look up the corresponding air flow in the compressor map.

This number should fit reasonably to air flow you calculate from AFR, and fuel flow derived injector size, injector duty cycle and differential fuel pressure.
The calculated air flow should also fit to your power level. If you burn much more fuel than you'd expect from the power the engine generates, you indeed might blow fuel out of the exhaust during overlap. At 5 to 8 PSI and high rpm this is usually not too big of a problem.

[jondee86]

Thanks for the comments. Due to space restrictions, I have two short
radius 90 deg bends after the throttle, so a little bit of resistance there
that I have to live with. However, I am aware that resistance will vary
as the square of velocity, and I always intended that the intake in front
of the throttle be as free-flowing as possible. I will be using an Apexi
500-A002 filter and 75 dia alloy tube with a single 90 deg bend.

http://www.visionr.com.au/images/site/products/306--500-A021.jpg

All going well this will get at least partially enclosed in a "cold air box" to
try and avoid sucking in too much warm air. Even if this only brings a small
gain, every little bit helps :)

I have no idea of the fuel consumption yet, as the car has not been drive
very many kilometres yet. But I'm getting hold of some 440cc/min DENSO
injectors, as the injector duty cycle hit 70% @ 5.16psi with the 365cc/min
injectors I am using at the moment.

Cheers... jondee86

[rotrex]

Again,
please measure the pressure after the throttle body before swapping air filters.
Only takes a few minutes and gets you a real idea how good or bad that or the new filter is.

You can literally route a silicone pipe through the passenger window. it is only a temporary measurement.
Hook up a analog gauge from eBay and record it with your cell phone. After than watching the gauge at WOT.

Just swapping and hoping things get better is the most expensive way to do things.
This is why most car dealers this days lost the art of properly diagnosing issues.
They read codes and swap expensive ECUs and other modules to only find that it was a single blown or corroded fuse socket in the end or water in spark plug holes.
They don't care and write you a 1000$ bill that should have been a $50 bill with common (car) sense.

[jondee86]

Please measure the pressure after the throttle body before swapping air filters.

Good suggestion :) Since the pressure loss of the ducting between the throttle
plate and the s/c inlet is fixed, I shall tap a fitting into the throttlebody in front
of the throttle plate. That way I can measure the pressure loss of the section
of ducting with filter that will be changed.

I have found cheap China made digital differential pressure meters on Aliexpress,
and this gives me a good reason to buy one, plus an infrared thermometer for
reading the temperature from each cylinder at the exhaust manifold.

Cheers... jondee86

[jondee86]

Existing tapping was for the evap system and read manifold pressure at low
throttle openings, so not suitable for reading intake pressure drop. I drilled and
tapped the throttle body for a new barb and used that. You can see it in this
pic of the new intake system...

http://i.imgur.com/amfZ8tt.jpg

Ran tests with the OEM filter and intake system and then with the new system.
The results turned out like this...

Intake Pressure Drop
Apexi Intake; 1.43kPa @ 7490rpm
OEM System: 5.91kPa @ 7480rpm

Engine Manifold Pressure
Apexi Intake: 143.3kPa @ 7490rpm
OEM System: 136.4kPa @ 7480rpm

So by reducing the intake pressure drop by 4.48kPa (0.65psi) I gained 6.9kPa
or exactly 1.0psi of boost :) I am more than happy with these results. When
working with small capacity engines like the 4AGE, every little bit counts !!!

Cheers... jondee86

[jondee86]

Getting some heat transfer into the alloy intake tube as it is resting on the
top radiator hose. Going to look for a different hose with a shape that will
allow it to clear the intake tube. Always something else :)

Cheers... jondee86

[rotrex]

1 PSI for free :-)
It looks like you are about half way there.
I'd have a close look at the tubing past the TB and any related edges, ridges or couplings to the intake of the SC.
The SC intake itself may also be restrictive due to tight corners or edges.


The same is true for the SC outlet to the intake manifold. I'd check for imperfections. The outlet is not as sensitive as the intake side regarding losses, but still.

I am sure with an other close look and a little fettling of the pipes and couplers you might find that other missing PSI or 2.

Heat transfer though this pipe is largely irrelevant except at prolonged idle and for at the first few seconds in after driving off in city traffic.

[jondee86]

Unfortunately I think that the ducting between the throttlebody and the
head is about as good as it is going to get. I did a bit of porting on the
short side radius where the discharge manifold meets the s/c mounting
plate and on the intake manifold where it meets the ducting. Removed
any weld that broke thru into the ducting and generally tried to get the
coupled joins lined up and as close together as practical.

Any further improvement will be hard to come by and at this stage I must
settle for what I have. The next few months will be dedicated to refining
my tune and ensuring that the engine operates safely under all conditions.

Cheers... jondee86

[jondee86]

Small update. Currently running on the 87mm dia pulley and making around
71-72kPa of boost while running very rich when getting into the new area
at the top of the map. Working on getting those tuned a bit better now.

My ECU can log in lambda but my WBO2 gauge can only indicate in AFR. So
to avoid confusing myself I shall continue working in AFR. Just drained the
water out of the tank and replaced it with 50/50 v/v methanol mix. So going
by your previous advice I shall continue to adjust the map to get the actual
recorded AFR as close as possible to my (conservative) target AFR's.

I shall probably look at installing the 75mm dia pulley in a few weeks, as
there was very little that required adjusting when I installed the 87mm dia
pulley in place of the 105mm dia. Once I have the 75mm dia pulley on the
engine I will look at getting the car on the dyne for some professional tuning
and to see if getting the ignition and fuel closer to optimum will help find a
bit more power and improve my gas mileage !!!

Cheers... jondee86

PS: Does a methanol and water mix always turn milky ?

[Richard L]

If it turns milky, the methanol is most probably contaminated with hydrocarbon. It will damage the aquamist components.

[RICE RACING]

PS: Does a methanol and water mix always turn milky ?

Not sure if its 'normal' but when I get 100% methanol then do the blend to water (WM50 by mass) it always went milky as well as quite hot, after some time it settles down and it goes totally clear.

I have some of the above that is in a long term test glass container and its crystal clear after 7 years (1/2 air in there the rest WM50), the components I have sitting in there are some valves, fittings, ATL fuel foam, and lines I use. I test the parts every so often and they all are working as brand new items.

I'll take a photo later if I remember and post it up.

[Richard L]

You are correct, it cleans after a few seconds. But it is doesn't, don't use it.

[jondee86]

The mix sat in my semi-opaque Pe tank overnight, and after reading Richard's
warning I went to drain and clean the tank. When I looked inside it looked like
the mix had settled, but there was a layer of "waxy platelets" floating on the
surfuce of the liquid.

If I can find a suitable glass jar I will conduct an experiment by mixing up a
sample and observing what happens over 24 hours. I do know that the meth
decanted clear and the "super distilled" drinking water I am using is also clear.
But when the two are mixed the colour change is virtually instantaneous.

Cheers... jondee86

[rotrex]

Chemist here.
Water and methanol mix in any ratio and are fully soluble in each other. It turns hot when you mix water and methanol. That is normal. There is no formation of any insuluble components unless the methanol or the water contains impurities.
As mention before, hydrocarbons, i.e. a wax, are the most likely contaminats in your case.
Shorter chained hydrocarbons from a petrol additive would be liquid at room temperature and float as droplets on the surface after some time.

I'd look for a better methanol.
I have used Efoy fuel cell methanoi.

For the indicated AFR of you wideband, tune for 12-12.5:1 (petrol equivalent) = lambda 0.8-0.85
If you run much richer you lose a lot of power.
You will also need sufficient ingnition advance. The mix burns a tad slower than pure petrol.

[RICE RACING]

The bit of advice missing on giving a target excess fuel level is to first consider knock limitation of the engine.

This is #1 in order BEFORE going for a arbitrary 'ideal' setting on fuel. There is no point targeting one thing if the engine is in 'pieces' ;)

IF you are needing WI in the first place this in most cases means you are knock limited and its very common knowledge that the tendency towards knock drops with 45% range of excess fuel along with WM50 being administered, using less greatly increase the level of knock seen and its pretty poor advice to give to say target a certain point, without mentioning the most important aspect IE: keeping the engine alive.

If you need to see graphs in terms people understand of real applications then I am happy to provide them, I can show you how marginal this is with various levels in Lambda terms from 0.70 to 0.73 to 0.76, it has a large effect.

[jondee86]

I made a test with about a 35% methanol to water mix in a clean container,
and it did get milky and it did get warm. Over ther first 24 hours there was no
apparent settling or change in colour, but after 48 hours I can see a clearer
layer forming at the bottom of the container. It is currently about 1/3rd of
the total sample.I shall continue to observe :)

After discussing the matter with the supplier of the methanol, it appears that
he recycled a container that had previously held another product. So he has
undertaken to replace the contaminated product with 20 litres decanted into
a fresh new container. I shall test a sample before using it to make sure it
remains clear after mixing.

Cheers... jondee86

[rotrex]

Personally, I'd return this methanol and discard the mix.
Any milkyness while mixing should disappear within say less than a minute. Mine never went milky, just some schlieren.
Milkyness means there is a phase separation, so droplets of one fluid in an other fluid. Methanol does not do that. Enthalpy of mixing and solubility too great. Your methanol contains a compount that is insoluble in water, but disolves in methanol.

Get fresh pure methanol, best in original containers. The residues, maybe even polymeric in nature, may clog your pump, valves and nozzles over time.
If the residue is soluble in pure methanol, I would flush the system and nozzles with pure fresh methanol and check the spray pattern of all jets before any driving.

Peter,
My engine was always knock limited. Knock happend before peak torque could be reached.
More mix allowed to gain more torque once ignition timing was optimized. I never went as far as finding the limit for the boost employed as mix consumption become too high for my track use and tank size. Power was good enough. Lotus Elise Mk1.
Only constant factor was that much richer than some 0.75 or some 11, give or take a bit, and achievable power of my set-up dropped under methanol water. 0.8 was fine, 0.85 was fine, 0.9 was fine. 0.7 it did not like as combustion went slow. I did not discriminate between initiation vs flame front speed. No ignition advance would compensate, so likely flame front speed.

[jondee86]

Get fresh pure methanol, best in original containers. The residues, maybe
even polymeric in nature, may clog your pump, valves and nozzles over
time. If the residue is soluble in pure methanol, I would flush the system
and nozzles with pure fresh methanol and check the spray pattern of all
jets before any driving.
Yes, the original delivery is being returned and will be replaced by a fresh
delivery in a new (unused) container. I shall test a sample before mixing any
for my spraying system. I shall also rinse my spray tank with pure methanol
as you suggest, to remove any residue remaining from the contaminated mix.

The car was not driven or the pump activated while the contaminated mix
was in the tank. I removed the tank and washed it thoroughly with water
after dumping the contents. So I think the system will be OK.

Cheers... jondee86

[jondee86]

The contaminated mix has cleared quite a bit but still looks "dirty" after
sitting for more than a week undisturbed. It seems uniform with no apparent
phase separation or sediment, but there is a thin oily/waxy film floating on
the surface that is evident when the container is held to the light.

Now the replacement 20 litres of methanol has turned up, and when tested
exhibited no milkiness whatsoever when mixed with water. The mixture
remained just as clear as the individual components. So onward and upwards :)

Cheers... jondee86

[jondee86]

So, question for the chemist :) How important or even necessary is it to
vent the water/methanol tank ? Will the methanol dissociate or come out
of solution with the water and evaporate if the tank has an open vent?

Motorsport regulations here stipulate that all fuel tanks must be fitted
with vents that terminate outside the cabin of the vehicle. But the regs
make no specific mention of methanol/water tanks and I would like to be
aware if there is any danger in having a non-vented tank.

Cheers... jondee86

[rotrex]

My Aquamist tank had a vent build into the tank cap.
Things won't separate in the liquid phase, but if you expose the tank to higher temperatures, e.g. parking in the sun in summer, the methanol will preferentially evaporate and build up a pressure in the tank. At worst, it might pop the tank.
So having a kind of a vent makes sense.

Also as mix is used, the fluid volume needs to be replaced by air in the tank. This has tom come from somewhere, hence a vent.
This is unless you run a bladder tank.

[jondee86]

OK... since I can't imagine that the generation of vapor or vacuum due to
usage being sudden or great, I will fit a small diameter (2-3mm dia) with a
fish tank filter/bubbler on the end.

https://i.imgur.com/2lA8yBp.jpg

Should prevent any dust or bugs getting into the tank :)

Cheers... jondee86

[jondee86]

I have made a few logs since starting to use the 50/50 v/v mix of methanol.
One thing I noticed was that on a short pull in second gear, the spray only
started to have an impact on IAT when the engine was almost at redline.
This is even though spray will have commenced early in the run. Most of the
cooling effect happened after I lifted off the throttle.

So I made a bit of a run in fifth gear...

https://i.imgur.com/IXaKZSl.png

This slows things down to the point that the cooling effect can be seen in
effect during the run. The s/c discharge temperature rising and the IAT
falling. Presently I calculate that I am injecting about 21% mix to fuel ratio,
but I do have one smaller nozzle available.

This brings me to my question... should I be aiming for the kind of temp
drop shown in the graph, or should I be aiming to simply hold the IAT more
or less steady ?? Currently I don't have any fuel or ignition compensations
geared to the IAT, but the tables are available in the ECU.

The engine is running too rich at the top end, and I am working on that at
the moment. And my apologies for the clunky graphs, but the laptop I use
for logging is old and can't keep up with the sample rate at high rpm's :(

Cheers... jondee86

[rotrex]

IAT sensors are rather slow. We talk 10-30 second response times for most.
The reason you see the air temp dropping is largely due to evaporation of excess methanol covering the walls of your intake wall and internal surfaces of the SC. They take time to cool down. The fluid films rapildy evaporates closing the throttle. IATs drop further.

I would aim for the amount of mix you need to maximiye torque. The IAT for this is only of secondary importance.

Due to the many turns, there won|t be too many fine droplets of water left.
I would therefore expect even better power with more methanol. It is more tolerant regarding droplet size.
Try 50:50 by weight. This is more like 100:80 methanol to water by volume.
It will do its thing even as entering the chambers as big droplets or streams.


IAT correction mapping work well for me starting with -1 deg at 50, -3 at 60 and -5 at 70C and up.
Base timing with cold IATs was optimized on the road. On track, the only way to get heat into a CC system, I then adapted the IAT correction table to prevent knock at these higher IATs. Worted a treat. The car was charge cooled, so it would maintain low IATs for single pull.

I found direct port to be the most efficient knock surpression for a given amount of injected mix. It also prvented me from bothering with cooling effects of methanol on the sensor. I got fluctuations from liquid droplets on the sensor that then translated into timing fluctuations.

[jondee86]

You must be using slow sensors. I "accidentally" hit WOT in first gear
yesterday, and my AUX Temp (same sensor as IAT) went from 70 to
100degC in just over 3 seconds according to my log.

Thanks for the input. As I don't have any practical way of knock detection
due to the amount of mechanical noise in the knock spectrum, I shall have
to leave ignition optimising to the dyno tuner. However, I will try the 50/50
w/w mix when I refill the tank. Currently I have 1 x 0.6 nozzle just after the
s/c discharge and 1 x 0.4B nozzle in the last bend before the intake manifold.

The intake manifold has an ugly rightangle bend which will be quite good at
inertial separation of droplets. My intention was that the 0.6 nozzle would
contribute to charge cooling and the 0.4 nozzle would get some droplets into
the combustion chamber.

If I can find a clear stretch of road, I want to do a WOT run using 2nd, 3rd
and 4th gears to see how high the s/c discharge temp will go. The s/c is
good for up to 150degC discharge temp according to the manufacturers.
Depending on what I discover, I may have to look at implementing a small
nozzle in the s/c intake manifold.

Cheers... jondee86

[rotrex]

Have you tried a external knock detector or indicator?
I first used a TurboXS knocklite. Worked well as a indicator.
I then installed a J&S Vampire knock controller. it also worked really well even on my engine running forged pistons. IT did a great job when I had a pump pressure issue. Engine was bogging and spitting flames, but the full 10 deg retard due to heavy knock kept the engine together.

There is chap in russia selling a reasonably priced knock indicator using a Bosch knock sensor. You'll be surprised how well such devices work.
Knock is also pretty audible in a car lacking sound insulation. It sounds like sparking. I can be heard even through loud engine noise.

https://www.ebay.de/itm/Knock-control-detection-module/292332868288?ssPageName=STRK%3AMEBIDX%3A IT&_trksid=p2055119.m1438.l2649

https://youtu.be/vztEn3wLi_c

[jondee86]

I have the factory knock sensor connected to my ECU and there is a
headphone jack on the ECU so I can listen directly to the knock signal. The
problem is that the engine is mechanically noisy and the s/c also makes gear
"rattle" noise in the same frequency band as the calculated engine knock
frequency. The knock signal is high at idle, at low rpm's and on deceleration
even when fuel cut is active.

At high rpm's the knock signal reduces to nearly zero, but when throttle is
reduced and the s/c gears become unloaded, the knock signal instantly
becomes high no matter what rpm's. So I need to be able to program the
background knock threshold levels, and discriminate between mechanical
noise and real knock.

Some of the high end knock monitoring can do most of this, like the PLEX
unit. Only it is a very expensive experiment if it does not work effectively !!

I have been considering trying a Bosch "donut" sensor in place of the
factory sensor, and locating it to a diffrent mounting position. Maybe that
will provide a better signal as I read that the wideband type of knock sensor
is sometimes more effective than the factory tuned type.

The knock light shown in your link is cheap enough to experiment with. But
this locally made product might be a better option...

https://i.imgur.com/tG6nmiw.jpg
http://www.nzefi.com/product/knocklink-g4-visual-detonation-warning/

A visual indication is probably the best option for me, as I will have to make
my own decision if it is knock or noise according to how I am driving at the
time the light comes on. Will have to decide on this soon :)

Cheers... jondee86

[rotrex]

I used both devices with the Bosch "donut" type of sensor bolted onto the block near the cylinders.

[jondee86]

I have been talking with one of our top tuning shops about this, and they
say that at the noise levels I am getting there is basically no way any kind
of detector will be able to separate real detonation from mechanical noise.

However, the human ear is very good at picking up the sound of real knock
even with high levels of background noise. So I will spend a bit of time with
headphones plugged into the jack provided on the ECU for listening to the
knock signal. If I can't provoke any knock by loading the engine at different
speeds, I'll leave optimising the ignition timing until the car goes on the dyno.

It makes sense that you either have to have a knock detection system that
you can rely on to only report genuine knock, or have no knock detection
at all and a good tune :)

Cheers... jondee86

[rotrex]

positive displacement superchargers can be very noisy. They oscillate in several vibrational modes of their rotors and cases at certain rpm ranges. This is one of the reasons OEM installed positive displacement blowers are mounted in a pretty stiff way to the block or are heavily braced.
A particular example that suffers form this are the EP§ Jackson racing superchargers on Honda K20 engines. IT pretty much renders the knock detection useless in certain rpm ranges, IIRC some 5000-6000 rpm.

The Rotrex SC runs very smooth and is rather quiet compared to a roots type blower. Although with a cone type of open air filter the intake sounds like a jet fighter, a sound similar (same) as turbocharger intake noise.

[jondee86]

positive displacement superchargers can be very noisy.
Yes. My s/c is a "short nose" version and the pulley is mounted directly onto
the end of the rotor shaft. It is common for "long nose" versions from many
manufacturers to have a coupling with an resilient element between the
pulley and the rotor shaft. The coupling absorbs a lot of the pulsations that
would otherwise be transmitted from the crankshaft to the rotor shaft. This
reduces (damps) harmonics and gear rattle. But it does introduce another
item requiring maintenance into the s/c installation.

You can clearly hear gear rattle from my s/c when warm at idle. But it does
start to make a satisfying whine when the engine gets up to 6000rpm and
the gears are no longer rattling :)

Cheers... jondee86

[jondee86]

I'm now running the smaller 75dia pulley on the s/c and using 50/50 w/w
methanol mix. I was out for a drive this afternoon and logged a short pull
in second gear that showed 195kPa @ 6000rpm. So with another 1800rpm
available, I think that 200kPa should be possible :)

The MAT and s/c discharge temps are quite a bit higher now with the 75mm
pulley. I can see as high as 95 MATand 115 degC s/c discharge cruising out
of boost at highway speeds (no injection). So I don't mind running a bit on
the rich side.

Car still runs very rich at WOT (into the 10's) and while I am dialling back
the gas to bring the AFR closer to 12:1 I am wondering if maybe I should go
back to the 0.5 nozzle I was using previously (0.6 currently fitted) ? Is there
any way for me to judge if I am injecting too much mix ?

Cheers... jondee86

[rotrex]

if you manage to achieve max torque before knock occurs, you could consider injecting less or run more boost.
if you you do not manage to get to MBT timing due to knock, you could inject more.
if you get misfires, you also might inject too much for you ignition system to cope with.
once you run it tad leaner than 10:1 while injecting the engine will wake up.

[jondee86]

OK. MBT tuning will have to wait until I get the car on a dyno. In the meantime
I will keep tweaking the fuel table to get the AFR while injecting into the low 11's.
Engine is running well with no sign of any hesitation under load. Did a 3rd gear pull
on the highway today and managed 199.6kPa when hitting the rev limiter. Or at
least I thought it was the limiter until I checked my log and found that I hit boost
cut at 197kPa. Didn't realise I had an overboost cut set, so maybe next time I will
manage to struggle past the magic 200kPa mark :)

Cheers... jondee86

[jondee86]

Had to enpty my tank and refill it today, and noticed that the methanol
smell did not seem as strong as when I first mixed this tank (50/50 w/w).
So is there a use by date for mix ? This tank would be about 6 weeks old
and I have only used about 1/3 of the 10 litre tank.

Cheers... jondee86

[rotrex]

no, unless you leave the reservoir cap off, a 50:50 (wt) mix will last pretty much for ever.
Do not use an aluminium tank for any methanol formulations unless it is coated on the inside with a appropriate coating.
Polymer containers work the best.

Air will slowly oxidises methanol in formaldehyde and formic acid. This in itself won't really bother your injection system.
Formic acid and methanol itself will slowly react into a ester with a distinct smell.
Methanol will react with formaldehyde forming hemiacetals and acetals. These have a distinct smell.
These chemical reactions of "aging" methanol exposed to air are very similar to what happens in raging spirits like whiskey or wine.
Their smell gets more complex over time as the above compounds are formed.
They all burn well and are only formed in rather small quantities with low reaction speeds given you start with high purity methanol.

Very high purity methanol has no intense distinct smell at all. It is almost odourless.

Ph.D chemist here. I have worked in the past with HPLC grade very high purity methanol and had a sniff. Natural curiosity.

The more "technical" grade it gets, the more off a soapy or booze like smell it gets.
As a mix with water, the oder intensity drops even more.

Some of the compounds you can smell are very volatile and easily vent off if you open the container. I'd not be surprised if you sometimes smell something and sometimes not.
Propanol and butanol have a smell that reminds me of technical grade methanol, especially when sprayed or hot.

This could depend of how long it was open before you exposed you nose to vapours, if the reservoir as cold or warm, direction of the wind....

When my car went into storage over winter, I have always drained the water methanol system.

[UCTURBO]

If anyone is curious, I mix my 50/50 with pure methanol and tap water and it stays 100% clear but when I add a small amount of methanol lube it goes cloudy and heats up quite a bit more. The only reason for the lube is I have some coated mild steel springs in my DP check valves that rust without it. It seems to stay mixed for as long as I need it to lol. Cheers

[jondee86]

All good then :) I am using a Pe container and I had to syphon the mix out
of the tank into other containers and that took a while. So any smell that had
accumulated in the tank would have dissipated while I was syphoning the mix out.

Got a few fine days now so I shall see if I can use up another couple of litres :)

Cheers... jondee86

[jondee86]

Bit of an update with progress. I re-tuned my fuel map to take into
account the enrichment provided by the WM injection. Took a while
and then I learned that I can set the end of (fuel) injection timing
with my ECU, and that threw the fuel map out again... so re-tuning
again at the moment.

Also tweaking the point at which WM injection starts when cruising.
At the moment it kicks in at around 110km/h and since I solved the
problem I had with the bypass valve not opening, my MAT's have
reduced to acceptable levels during around town and cruise. There
is still some work to be done but the car is running well and I have
been able to enter a few local events like this one...
https://www.youtube.com/watch?v=8pSyUqAevKA

I have had to up the idle to 1250rpm to get above the annoying gear
rattle from the s/c. Othere than that this setup just keeps getting
better and better. Not overly fast, but oooooohhhh... that midrange :D

Cheers... jondee86

[jondee86]

I am attempting to resolve a tip-in misfire that can be seen here...

https://i.imgur.com/1iXaOtz.png

When starting to accelerate, TPS movement invokes some "throttle pump"
enrichment, and the increase in injector duty cycle triggers the WM injection.
This appears to cause a rich misfire that shows as a lean spike in the AFR.
My WM injection initiates at around 10% IDC as in the absence of an inter-
cooler, I use WM for charge cooling once the car reaches around 110 kph
while cruising.

I have a number of adjustments for both fuel and WM injection, and to help
with getting the balance right, it would be nice to be able to log the actual
onset of WM injection and the rate of injection.

From reading the HFS-2/3.1 manual I gain the impression that if I use the
Green Harness I should be able to output a 0-5V signal from my HFS-3 to
a spare 0-5V analog input on my ECU.

EDIT: OK, looks like the yellow wire on the green (USER) plug should give me
the information I require, so I will go with that.

Can someone please confirm that my understanding on this is correct ?

Cheers... jondee86

[jondee86]

Got the flowmeter output in my logs now, and it shows me when the WM
injection triggers and how the flow varies with rpm. Now I can fiddle with
the various trims available to adjust bothe the trigger point and injection
rate to try and get a better match to my fuel injection IDC graph.

Finally managed to find a decent OEM bypass valve (off a turbo Legacy)
that I could modify to fit my Bosch based bypass pipework layout. Should
allow a bit more air to flow when out of boost, which should in turn help
my out of boost cruising manifold air temps. Will be testing this in the next
couple of days to se if it performs as expected.

Cheers... jondee86

[jondee86]

New bypass valve is up and running...

https://i.imgur.com/xAvn9VG.jpg

Works great and seems to pass a bit more air than the old system, so I am
having to re-tune my VE map below 85kPa. Engine pulls hard all the way to
my 8000rpm redline with virtually no dropoff in acceleration. I am considering
getting a 65dia s/c pulley to replace the 75dia pulley I have at the moment.
This would mean dropping my redline to 7000rpm to avoid overspeeding the
s/c but I don't see this as a problem. It should help the engine live longer.

Currently I am chasing the misfire that I get when I accelerate hard in any
gear from around 2500rpm. It only happens once about a second after I hit
the throttle, and I have not noticed any misfiring further up the rpm range.

Bearing in mind the earlier comment made by rotrex about spray killing the
spark, I did wonder if that might be the problem. Hard acceleration at around
2500rpm invokes extra enrichment and also triggers the W/M spray. But when
I switched off the W/M spray the misfire still occured, so now I am adjusting
the enrichment tables to see if I can find an answer.

Cheers... jondee86