Tchaps

Thanks mate.

Certainly a lot more involved in terms of repackaging the engine bay than my previous hotside setup.

Hopefully the throttle response and generally larger capacity supercharger makes it worth while.

Tchaps

It’s in! I have driven it!

I had it idling without the chargecooler and radiator on Thursday then on Saturday morning I’m on it again to sort a few issues.

I noticed with the car up on stands the steering had become notchy. Luckily this turned out to be a grommet in the bulkhead snagging. Must have knocked it putting in the clutch pedal. Quick poke with a screwdriver and all was well.

Not so much of an easy fix was the issue that the clutch was not disengaging. I had put the clutch plate in back to front! So engine and box out again. Several new swear words were learned.

This mornings battle was with leaks on the charge cooler system. The pre radiator was in, plumbed, filled, inspected, drained, removed, welded, fitted again no less than six times.

I had also put the vacuum hose for the brake booster on backwards meaning I had no power brakes. Fixed.

Issues on the drive were ecu coolant temp was reading 78deg (82 thermostat) but my water temp gauge was reading about 105. I think I have mixed up the water and oil temp sender wiring.

The diff sounded horrendous. Going straight no issue but turning the clutches were chattering very viciously with a lot of noise. A bit of research seems to point towards me needing a friction modifier additive in the oil. I have ordered and will collect this tomorrow.

My oil feed restriction fitting I made is also slightly weeping. I thought it was on the thread into the sc but after further investigation I think there must be a tiny pinhole in the weld between the restrictor and -4 adapter. I will reweld this tomorrow to see if it helps.

A few stupid mistakes and a couple of small issues. Not too bad really bearing in mind the amount of things changed or rewired.

Also the engine mounts do not seem to be an issue, idles at 900rpm and if anything is smoother than before. Lucky draw I guess.

On the drive it felt good, it was only short and on city roads but it behaved and had great response. I gave it a tiny 75% poke and saw 10psi. I am hoping these baby pulleys will get me 14psi / 230whp. Dyno is on Tuesday so we will see. I will get some pics and videos up tomorrow

Here is a link to a quick video of it idling for the first time:

You Tube

And a couple of pics

Ted75zcar

iTrader: (2)

sweet!

Tchaps

Well well well what a rollercoaster!

So I took the car out for a test drive Monday evening just to check everything out. All good for about 15-20mins, until I go to accelerate and the car hesitates massively. I try again and the same happens then it catches and is normal again.

However the engine and supercharger don't sound the same. I limp it home and it sounds like it is missing a cylinder. Not misfiring but just not all there.

I was very worried I had damaged the supercharger or the engine.

I was pretty sure it was spark, maybe the wiring, but it is late and I can't check it out.

So I was up at 4.30am in the morning of dyno day and when I pull plug no.4 it is a lot whiter than any of the other cylinders. It looks like it has run lean, which ties in with the fact my fuel autotune I was running had added about 20% fuel to the cruise areas I limped home on last night. I was very nervous, had I caused some permanent damage? I scoped the bore and everything seems ok, checked valve clearance to try see if I had burned a valve. Luckily everything was clear. I went though my injector harness and resolderd a couple of joints, cleaned the injector no4. When I refitted I swapped injector 4 into cyl 1 to try to help diagnose if it happened again. It was 8.30 and I fired it up and everything was back to normal.


Eek


Hop in the car for the hours drive to the dyno and everything seems ok. The oil temps are hotter - running at 115degC but it doesn't seem to matter what speed or road type you are on, it just sticks at 115. I will possibly need a larger cooler to give myself headroom on track but we will see.

Got to the dyno without any further issue. We go through the ECU and change some of the wiring (injector grouping and grounds). We had a little difficulty getting the car to idle dead steady. There was a small air leak on the injector seals which we fixed but there still seems to be a very small leak on the manifold I made around cyl 3. Only tiny but just means it idles at about 13.5 AFR to be happy. I will pull it off and change the inlet manifold gasket (which was the one from before this install - possible laziness on my side) or maybe a pin hole.

Anyway on to the main event - The car was making 10psi of boost, so dissapointingly I was expecting 210whp, down from my best hope of 230whp at 14psi. First pull it made 236whp! In the end it made 252.4whp / 282.8hp at the crank This is on the baby pulleys - spinning the supercharger up to about 15550rpm at 7100rpm. All this at 10psi - which I am astounded by. My M45 setup made 180-190whp at 8.5! The flow with this setup has really suprised me, and this is also with 1.5" exhaust primaries.




The big boy pulleys will spin the sc up to 17750 rpm - A bit of quick maths says 288whp. It is likely that the exhaust manifold will become a restriction at this point, and also the output temps of the supercharger climb by around 30degC between 15000 and 18000 rpm so that may reduce the efficiency. However from the dyno plot attached the power 'curve' has not crested yet. By gearing up the charger this should hopefully mean that the actual peak output of the charger can be achieved within the 7100rpm limit and move peak torque lower in the band, this could boost final figures above the direct relationship I have calculated from.

We shall see, my tuner Dale thinks we could hit 300whp.

Numbers aside the car is fantastic to drive. The pickup is instant and smooth. The power delivery is a lot more refined than my m45 setup. That used to come in with a punch (I think because of the dual tbs) but then be absolutely linear up to about 6k then start to roll over. The Autorotor is smoother to start but as the revs build it pulls harder and harder until 6k -7k where it feels like a spaceship even on the small pulleys.

From not even thinking I was going to make it to the dyno to a result that completely surpassed my expectations by the end of the day really boosted the elation. I could not be happier with it.

I have to pull the diff at the weekend to change the oil and recheck clearances, then hopefully the party pulleys will arrive before the track day.

I will share more data as I get it from logs but I think it was running about 16deg at 10psi - I am going to download the map this afternoon and take a look.

I have some videos that I am posting on youtube of the dyno. I will post the links once I have them. The sound on the videos is a lttle underwhelming if Im honest - it sounded absolutely mental in the dyno room in person!

Tchaps

Links to a couple of dyno videos:

Mid range tuning:


Not the dyno pull with the final numbers but I think 3rd from last. Still made 250whp:


Enjoy!

RunninOnEmpty

iTrader: (5)

Amazing work! It looks practically off the shelf! Work like this reminds me where I stand on the DIY hierarchy.

Tchaps

Ha ha ha thanks mate appreciate it.

Oddly that GIF is exactly the process of stripping the diff down, maybe with a bit more cursing

Since the mapping I have been working through a few issues.

I took it for a drive on the evening after mapping and it dropped a cylinder again. Same symptoms as the night before mapping. I nursed the car back home and pulled the plugs. Plug 4 was still whiter than the rest but cylinder 3 was wet with fuel. A bit of diagnosis work pulling leads off spark plugs confirmed that cylinder 3 was the problem cylinder, and the issue was no spark. A bit more root causing, swapping HT leads, spark plugs and coils onto cyl 3 wiring confirmed that it was infact the wiring to cyl 3 that was the issue. As I am wasted spark and cyl 2 was fine it had to be a dodgy connection in my Mx5--> LS coil sub loom. I got new pre made pigtails and used soldered joins instead of crimps and loaded back into the car. I also did a better job of wire routing than my first attempt. The car has not dropped a cylinder since so I think I got to the bottom of it. When I head to the dyno next time Dale the tuner is going to upgrade the drivers in the ECU also just to be safe.

I replaced the inlet manifold gasket to try to get rid of some of the vacuum leaks I was seeing around the ports. It now idles far smoother, and also spark plug 4 is the same colour as the rest now. I think there was a reasonable leak on that port join causing the cylinder to idle lean. The old gasket was fairly mullered in that area.

Next up was the rear seal on the gearbox was absolutely chucking oil down the underside of the car. The leak was so bad that it was making the diff look like it had a leak! When I removed the old seal it wasn't OEM but it did look in ok condition. I was concerned that I was missing a trick and that maybe there was a 6 speed specific prop shaft or seal (there isn't) because I couldn't understand how so much il was getting past a seal that looked ok. I replaced the seal with a new OEM part, took it for a drive and the leak has stopped, now completely dry. Should have done it before I put the box in the car but there we go.

I replaced the diff oil now the LSD is worn in enough. When I was doing this I noticed that the LH half shaft was having a go at the diff casing. I have the 2 piece 1.8 halfshafts on my car. They have a seal protector just as they mate into the diff, I assume to stop grit and road grime being able to get directly on the seal. On the LH side it looked like the shaft was pushed in too far, and this seal protector was right up against the diff casing, wearing it away. The RH side was fine, a small gap between the protector and casing when pushed in as far as possible, as it should be. I pulled the LH shaft and the circlip was still inplace. I removed the protector from the shaft and plugged it back in. It felt normal, I could feel the circlip engage in the diff and it had a normal amount of movement in and out when engaged, it just seems to be that this engaged position is slightly more inboard than it was / should be. I will run it without the protector for the moment while I talk to the shop I got the diff from if they have seen this before. I will swap the LH and RH halfshafts at some point just to see if it makes a difference but I just wanted to get the car back on all 4 wheels for the track on Saturday. The seal looks like it has had a bit of abuse from the protector but it is still intact and not leaking. I will replace it when I do a more thorough investigation.


LH side can be hard up against the case, grinding against it.

RH always has a small gap

You can see where it has had a bit of a go at the case and seal. Seem to have caught it in time luckily.


I should get my upgraded pulleys today, so should be able to fit them and go for a run this evening to check the fuelling and timing are ok.

Tchaps

Haven't updated this thread in a little while, apologies. Got some good / interesting updates.

I got the pulleys, they fit fantastic. They did the hard anodising the wrong colour (I asked for black, they did them clear) but to be honest the natural colour has grown on me and I think it looks really good.


ATI crank overlay, 160mm 8 rib

64mm 8rib supercharger pulley

All fitted

Then on to the trackday:

First of all, it survived! It performed fantastically all day without any supercharger / ignition / tuning related issues. I was doing short runs in it to build confidence. The track is quite short (1.6miles) but has 3 short straights, with a mix of fast and slower turns. I was keeping runs to 5 or 6 laps at a time. The supercharger intake manifold and support has held it all up bumping over curbs without issue.

The increase in power has certainly made driving the car on track in the dry a more intense experience. It has been a while since I had been to this track, and since my last trip I have also upped the tyre grip levels. I found that the medium speed corners I could take an entire gear higher, despite with the 3.6 rear end 1-5 on my old 5 speed trans and 1-5 on the new 6 speed being roughly equivalent. The additional power meant if I took the corner in the old lower gear I would run out of rpm before corner exit. Using the taller gear allowed me to ride the torque and kept the car far more settled.

The diff was very impressive out of slower corners. I did not suffer traction issues and the car did not feel over powered which I found surprising. You can certainly feel the diff locking and it gives a very stable feeling accelerating out the corner, and the ability to get on the power far earlier than previously. It has certainly made up for the horrible noises it makes when parking in a car park!

I need to sort out some data logging for when on track. I have a tablet with shadow dash on it, but couldn't get it all setup in time. From my gauges both engine coolant and charge cooler coolant temps stayed right where they should. Temp on the day was mid 20s (deg C). Engine coolant was stable at around 85-90degC. Charge cooler temps just about managed to get on the gauge scale, which starts at 40deg C. I realise this only tells half the story in terms of charge temps, but it atleast shows that the flow rates / size of pre rad / volume of coolant is adequately sized to remove the amount of heat that the 2x laminova cores are able to remove from the charge. When I sort the data logging I will take a look at what the IAT temps are with periods of prolonged high rpm / throttle.

The breather system also seems to be working well as far as I can tell. There isn't any oil in the catch can or intake tract so the position of vent and extra baffling in the valve cover seems to be doing the job, keeping the oil inside the engine. The oil is also still clean so I hope this means that the blowby vapours are managing to be vented out of the engine. I know from on the road that the crankcase remains at ambient pressure when under load. Its a bit hard to tell exactly how it is working but all the indicators so far are positive. The engine does not seem to have used any oil over the whole trip, 200 road miles, around 60 track miles. I knew the engine was tight from my previous M45 set up, but the new Autorotor uses the engine oil feed and I did not know the condition of the seals in it for sure. My engine also makes a lot of vacuum because of the short duration cams, which does not help the seals in keeping the oil in the gearcase from getting into the rotor housing when under vacuum.

There were some issues though:

1) The engine oil needs more cooling. This was not surprising, I knew from just driving about on the road that the SC had added about 10deg C to temps, before the new pulleys. On the drive over to the track oil temps were stable at 115deg C which is not an issue at all for the 10W50 I run. On track however temperatures were climbing up to 135degC - 140degC, and also not levelling out. My run lengths were dictated by the oil temperatures. On this short track I didn't mind so much, the purpose of the trip was to shakedown the new setup, but on some of the longer tracks I like to visit, the limitation of a 10 mile stint would get a little frustrating. The 10W50 I use is rated for constant temps of 125degC, max periodic temps of 150degC. I would really like to get the oil temps down to levelling out at 120degC on track, so looks like I need to package a larger oil cooler than the small 10row currently on the car.

2) The 149K 6 speed box is a little tricky. I kept missing 4th gear when changing down from 5th. The gate for 4th seems to be very narrow, and angled slightly towards me rather than straight up. I got used to it a bit more by the end of the day but it certainly takes a bit more conscious effort than my old 5 speed. It does seem to have handled the 300 or so hp abuse all day so fair play to it for that. The in cabin clutch master assembly worked great all day.

3) This one is completely my fault! I locked my front wheels up pretty badly as I was dialling in my brake bias (I have an in cabin adjuster attached to an adjustable prop valve in the engine bay) and flatspotted my front tyres. It was actually the first time in the dry on this pad setup (DS1.11 fronts in my 4 pot Wilwoods, PBS procomps in the rear on sport calipers) and it worked well when I had it setup but unfortunately I had already done the damage.

Not the end of the world as I had swapped the tyres front to back before the event, so the fronts were low on tread anyway and I would be replacing them for the next track adventure. Just a bit annoying that I will have to replace them this month now if I want to drive the car as the judder it causes is pretty unpleasant over 60mph.

In summary though I am very happy with how it all worked and drives, and that the supercharger did not melt and throw itself through my engine. Looking forward to getting the dyno session and having a play with the cam timing.

Tchaps

Following on from the trackday I had another session on the dyno organised to have a final tune on the new pulleys and have a play with the cams.

The expectation from the new pulleys, bearing in mind they had increased manifold pressure from 10psi to 13 / 13.5psi was 280whp / 320hp.

First run, just from the safe tune we put in for the higher pressure areas we got basically exactly the same peak numbers as at 10psi.

Advancing the timing yielded no gains to peak figures, we put 4 degrees into it without change. It also did not run into a knock condition either.

In the end we got 255whp and 285hp peak figures - quite disappointing and a little deflating.

We also tried moving the cams around, advancing / retarding one and both cams together. No power differences were found, infact the original position of 114deg max lift on both yielded the lowest kpa (slightly) and so the lowest restriction, so we just left them there.

After the session I analysed the figures more and we have picked up significant midrange. Excuse the home made graph, the dyno print outs are individual with no other way to overlay, in wheelspeed and with no torque curve.


whp/wtq = old pulleys whp2/wtq2 = new pulleys

Torque at higher rpms is decreasing far faster than previously and limiting top end. Got a few theories that need investigating:

> The supercharger efficiency map does not show a significant drop off of efficiency, and mass flow rate continues to increase linearly as rpms increase. This ties in nicely with the linear and accurately predicted rise in manifold pressure between the pulleys. Really the only trend that differs significantly looking at the supercharger in isolation is outlet temps go from 100degc @ 15krpm to 130degc @ 18krpm.

> Removing the intake tract from the supercharger on the dyno and running with an open tb only gained us 2hp at the top end. I think this is a pretty good result to be honest.

> IATs were certainly hotter than before. It was a very hot day, about 32deg or so. We saw IATs in the 60degC + range. This is certainly higher than I would like and I need to look into it. The 2x Laminova cores are only rated for around 260hp so we are pushing them. The hot weather has a compounding effect on the dyno, ambient air intake temps are higher even before the supercharger adds its heat, the charge cooler coolant can only be ambient temperature or above and there is reduced air flow to cool the charge coolant. I have since the dyno session done some datalogging on the road and on two gear pulls in equally hot weather I have seen IATs up to 57degC. I think the chargecooler needs a bit of help so I am looking into water injection. Seems to be the most cost effective solution. Ideally I would add another core to the chargecooler but that would mean basically making a new housing, could be done but I think far more expensive than spraying some water.

I think there could be improvements made on the intake side but given that IATs were not massively higher than with the old pulleys and the manifold pressure at the ports had increased proportionally as expected I don't think this explains the top end drop off.

My suspicion is that I have just reached the limit of flow for my exhaust system, with its stock sized 1.5" primaries, 2.5" downpipe / exhaust to a quiet muffler.

I am conscious there is a lot of 'I think' and 'my suspicion' in the above, so what really needs to happen is some data gathering.

The intention is to measure pressures throughout the system, all through the intake and the exhaust to see if the data gathered points towards a bottleneck.

I have already taken some pressure data from the sc inlet, post tb and found a differential pressure of -0.7psi from ambient at WOT 7krpm, which would certainly be worth finding the source and addressing it.

I have got hold of some fittings to measure back pressure in the exhaust at the port and the muffler entrance.

I had another issue with the ignition, I think the intake broke my HT lead connectors (it rests on them) so the car is off the road waiting for new leads, but hopefully by the end of the week it will be running again and data acquisition can continue.

What I am aiming for is to get over 300crank hp, with IATs under control, down to maybe max +10 deg C ambient for reliability on track.

3rdCarMX5

iTrader: (1)

Thanks again for the update. This is a great build. I am starting to get my version of this on the road again. I am currently going through the tune again N/A before bolting the SC back up. I plan to run flex fuel and the smaller pulley (bought 2.25" from kenne bell before they ran out of stock). The bigger pulley I have was 7-8 psi from memory, I think the smaller one is 12.5 psi.

Would you be willing to share your tune with me (I have an MS3x)? I plan to head to the dyno in a couple months, but it would be great to learn from what you've done. When I last had the car together, I was really struggling with my Accel pump settings. The SC really moved a lot of air when the throttle blade opened, which I think was causing transient knock.

It would also be great to share idle settings, as I found things a little more temperamental with the load from the SC.

Tchaps

No problem mate see attached.

I can't really take any plaudits on the tuning, Dale my tuner sorted it for me but talked me through some interesting points on what he was doing.

Spark is calculated using hybrid timing - the primary ign load is TPS % (alpha N). We then use kpa as a secondary load to pull timing when at 100+kpa. To be honest I found this odd but the reasoning is that due to the extremely short runners and small plenum volume, with a supercharger that expels air in big pulses the low load and idle pressure signals would not be the most stable and could give tuning issues. Fuel load is speed density. Idle control is primarily managed via spark advance, rather than the idle control valve. The valve is only on open loop warmup control. Personally I have found that the spark advance management of the idle is far more quick to react than my old set up using closed loop idle valve control without any spark adjustment. When the car is revved or when pulling up at a junction the revs settle straight away into idle without significant droop, with varied electrical loads on it.

Accel enrichment is still a bit of a mystery to me but you will see how my car is set up in the tune file.

Seems to work well so far.

Good luck with getting the Kenne Bell back on it and on the road. Wish we had E85 over here!

3rdCarMX5

iTrader: (1)

Thanks for sharing! I will have to take a look. I have an air to air intercooler, and a standard intake manifold. So I should have a fairly uniform MAP signal (and therefore plan to use a the standard RPMxMAP table). But the total timing could look similar If I can account for MAP to RPM conversion.

You are making me think that a blow off valve may be worth investigating. I have the bypass valve, which should work to vent the positive pressure back into the SC intake- but I will look into testing the pressure in the intercooler.

Thinking out loud, it seems strange to me that you added a lot more air but ended up in almost the same spot... I think that you may have a flow restriction through the intercooler (makes the most sense, or the tune is pulling timing based on an increased IAT). Do you have a port to check pre-intercooler boost pressures?

Tchaps

Just bare in mind spark table 2 is subtractive from spark table 1 to get total timing - it looks very very odd otherwise! With your A/A you won't have those same potential issues, plenty of volume to damp the pulses. Not even sure if it was going to be an issue on mine tbh. I drove it to the mapping completely on speed density and it worked fine. The car was far better driveability wise after mapping and the change to hybrid (un surprisingly), but far more than just the spark calculation was changed.

The pressure take off is post chargecooler, and because pressure increased proportionally as predicted with the increase in supercharger speed the additional air seems to get to the ports. I think if there was a restriction in the charge cooler the pressure at this position would not have increased as expected, I would have maybe only seen 11 or 12 psi at the ports.

My maths is @ 7000rom

Small pulleys: 15333rpm sc = (14.7+10)psi

Big pulleys: 17500rpm sc.

Calculation for manifold pressure with big pulleys: (((14.7+10)/15333)*17500)-14.7= 13.5psi

It is a simplification, but should hold reasonably true and current manifold pressure under WOT at 7000rpm is bang on 13.5psi

I might be wrong, but I hope I'm not, taking a pressure or temp reading post sc pre chargecooler will be a total PITA, there isn't really a good place to drill a hole with enough room for a take off or sensor. And if I do find an issue there I'm not really sure what I can do about it!

I am hoping that my exhaust investigations yield some clues, because I really want to have a good reason to build a new manifold!

I spoke to Dennis again about the results I got from the dyno, and he did tell me that he had encountered a similar set of circumstances, i.e spinning the sc faster, within efficiency range but yielding no additional power. In that case it was the stock catalyst / exhaust corking the setup.

3rdCarMX5

iTrader: (1)

I did some math, which I hope was correct and used a lot of assumptions, and with the delta in IAT (30c) you should have still seen a 19% increase in molecules of air going from 10 to 13.5psi. Did you see a 19% gain anywhere in the rev range? If it just ran out of efficiency, I would have expected a bigger gain in the middle (you could try to find the exact new engine rpm that matched your previous SC pulley rpm and compare the % difference). It is strange that you hit the exact spot you started. I know that the guys who are running more than 300hp were preferring a much larger header diameter, a popular one was made by BOIG (hopefully that helps you in the design process). You could also measure the exhaust back pressure between the two pulleys, if you wanted to run an experiment.

http://www.baautox.com/forum/index.p...th=10971&rid=0

Tchaps

Thanks for crunching the numbers from an AIT standpoint, I hadn’t done that.

No, nowhere near 19%. Around 10% in the midrange.

I’ll have a look at engine torque at the same sc rpm, would be interesting to compare.

Ive seen the Boig headers and I will certainly be taking some cues from them. i believe they are 1 7/8 diameter primaries - just seen they are from that link you sent

Tchaps

The car works again! Back on the road on 4 cylinders once again. Replacing the HT leads has done the trick, hopefully the new position of the coils will avoid the issue happening again.

Went out and did some more testing this morning, with my pressure sensor hooked up to the intake tract, about half way between the throttle body and air filter.

Conditions were about 17degC, so a lot cooler than previous logs and ambient pressure was reading about -0.05psi.

Cruising AITs were around 34degC, so around 13deg lower than previous logs, taken at 27degC ambient temps. Not a particularly surprising result.

I did a longer pull, 3rd to 4th, brake hard, 3rd to 4th again. IATs topped out at 44degC, the sensor saw this max temp about 5s after the end of the second 4th gear acceleration. So the sensor is seeing a max temp of 27deg over ambient, 10deg over cruising IAT. This result is not very different at all from my previous logs, just the temperatures shifted down by the difference in ambient temperatures.

I am also fairly confident I am seeing somewhere very close to the actual IAT temperature, taking into account the lag of the sensor. On the first 3rd/4th pull IATs rise sharply on the log, going from 34deg to 43deg in about 7seconds, about 1.5secs after the throttle is released to brake hard. When throttle is applied again for the second 3rd/4th pull I actually see a reduction in IAT by 1deg to 42deg, until 4th gear when the IAT slowly rises again to peak at 44deg 5seconds after the run. I do not see a second spike upwards in IAT during the second pull, which suggests the sensor is up to temp with the intake charge by this point.

44deg is cool enough for timing not to be pulled by the ECU and I could feel the difference. I do wonder whether the pull on the dyno on the hot day was long enough for timing to start to be pulled. It would go a long way to explaining the top end power reduction, but I don't have the log so I don't know for sure.


First 3rd & 4th pull datalog with rpm, boost, MAT/AIT and intake tract pressure

Second 3rd and partial 4th pull (ran out of road) straight after hard brake from first pull. Compare the IAT/MAT gain.

I still think a bit of water injection pre sc is a really good option to take a few degrees out of the intake charge so that on track on a hot day I don't run at 60-65degC intake temps. I was swaying towards the Aquamist HFS3 system, as it uses a PWM valve to meter water with respect to injector duty, rather than other systems that are either on or off, or only modulate flow in reference to boost (but not rpm also). It has a built in failsafe and signal for the ECU. But I am now thinking from the results above that it might be a bit over the top for what I need, in that I am not necessarily going to run a different map for WI, it will just keep IATs low so the ECU doesn't pull timing from the normal map. Need to have a think and a bit of a price comparison.

Something else to consider is that as the WI will reduce the temperature differential across the intercooler, and therefore its efficiency, the end IATs may not be dramatically different. For example currently at around 20degC ambient at WOT / 7000rpm sc outlet temps are 130degC, intercooler coolant 20-25degC. The intercooler is currently reducing these to 44degC, a pretty impressive IMO 86degC drop. If WI drops the sc outlet temp to say 60degC, the temperature differential between the charge cooler coolant and charge is only 35-40degC, down from 105-110degC. This will significantly reduce the temperature drop of the air charge within the intercooler, not sure by how much but I doubt it is linear as charge temps and charge coolant temperatures get close to one another.

Likely to be a large amount of experimentation required to optimise the IAT aspect of WI through my intercooler.

The results from the pressure sensor in the intake tract show I have a pressure drop of around -0.63psi from ambient at 7krpm. Again the results show a reduction in tract pressure as rpm increases and the relationship is essentially linear.

I am very pleased with this result. It shows that my fabricated sc inlet manifold is only costing me maybe 0.1psi. Next time I have it off the car I will sand a larger radius on the join as the pipe meets the sc flange at the top of the inlet as it is a little harsh at the moment but I'm not going to loose any sleep over it. Next test will be to take a reading from ahead of the air filter to see if the filter / air box is causing the small restriction.

Tomorrow I am hoping to weld on the exhaust fittings to begin testing for backpressure.

Tchaps

Two items to report back in to the thread.

> Tested for exhaust backpressure about halfway along cylinder 1s primary. They look so small once I removed the wrap! Measured the diameter at 40.5mm. I drilled a small 3mm hole in the primary, then welded a 1/8npt boss over the top and screwed a 1/8npt to -3 adapter into it. Slipped the flare nut on some 3/16 brake line and flared the end and screwed it up to the adapter on the manifold. I coiled the brake line a few times in the engine bay then fed it out of the bay, through the passenger window and attached the end to an analog pressure gauge. The length of brake pipe worked well. All of the pipe out of the engine bay was cold to the touch at the end of the 20min or so drive. I have attached some pics of the setup.





It worked very well. Got the car up to temp and gave it some beans. Backpressure at the manifold is as follows:

5krpm - 3psi
6500krpm ish - 6.5-7psi

Didn't register any back pressure below 4krpm, the pressure steps up quite rapidly in the 5k+ rpm range.

Conclusions from the above is that 7psi of backpressure at redline strikes me as quite high and something that should be reduced.

I need to test the exhaust system at the muffler entrance to see what the pressures are there also and compare to see if I can map pressure increase through the whole system.

TBH though I think I am going to build a manifold anyway, because I would like the fabrication challenge. It is nice that there seems to be some kind of scientific justification for that though

Tchaps

The second item is I have been having a read up and research on intercooler efficiency.

The particular question I needed answering was how lowering the air charge temperature entering the intercooler effects the heat transfer capability and ultimately intercooler outlet temperature. I understand the the heat transfer capability increases as the temperature differential between charge temperature and cooling medium increases, but I did not know if it was a linear relationship or not. I probably should have listened more in my old thermodynamic lectures.

I found this really helpful research paper on intercooler efficiency.

https://pdfs.semanticscholar.org/ba1...1164975cbd.pdf

One of the tests they did experimented with altering the charge air temperature at a constant mass flow rate and plotting it against Q, or heat transfer rate. They tested this at different cooling air mass flow rates (not of so much interest to me). See page 12, figure 11 on the document I linked.

What is of interest is that the relationship between heat transfer rate or Q has a linear relationship to charge air temperature entering the intercooler.

This is excellent news as it means that the more I cool the charge in the supercharger with water injection the better. The intercooler heat transfer rate may reduce with a lower charge temperature, but ultimately the air exiting the intercooler will always be cooler. The only limitation to this is if the supercharger charge temp is colder than the intercooler. My further calculations show I really don't think this is going to become a problem.

Using some of the data I have gathered from both logs and the supercharger efficiency map I took it further to look at what intercooler outlet / IATs I think I could achieve. First up Q for my intercooler has to be calculated.

Q=m*Cp*dT
m=mass flow rate - works out to be 0.304Kg/s @ WOT/17,500rpm/20degC ambient temps from sc efficiency map
Cp = Specific heat capacity of air at 130degC (sc outlet temp at 17,500rpm/20degC ambient temps from sc efficiency map and 0% humidity.
dT = Delta temperature across intercooler - 130degC (from SC ef map - 44degC from data logging at approx 20degC - ok it was 17degC but give me a break!)

This works out to be 25.884, not sure what the units are. Not amazingly useful on its own but with a few educated assumptions I think it could be.

So if we take the findings so far that with charge temps above cooling medium temps the relationship between Q and air charge temps is linear, and the assumption that if charge air temp = cooling medium temp the Q is 0, i.e no heat is transferred if charge temps and the intercooler are both the same temperature then we can build a graph of Q against different charge temps for my intercooler. The assumption I have made so far for intercooler temperature is that it is 5degC hotter than ambient.

The trendline for Q can be represented by an equation against charge temp and intercooler outlet temperature. Once the equation for Q is known as a function of charge temperature you can feed it back into the original equation for Q but rearrange to find the delta T, and therefore the new intercooler outlet temperature for a given charge temperature. I have attached a graph that for my intercooler at 20degC ambient temps, 25degC intercooler temperature.

The equation for Q changes with different ambient temperatures because of the small change in Cp and the different intercooler, sc outlet and data logged outlet temps, but if you have known values for one point on the Q against charge temp graph the new Q trendline can be calculated. For instance I am most interested in how to reduce intercooler outlet temperatures at high ambient temps. I have data from around 30degC. I got intercooler outlet temperatures of 60degC, I can take a good guess at charge temperature from the supercharger - +a little bit more than the difference in ambient temperatures, say 143degC @ 30degC up from 130degC @ 20degC.

I built this into a spreadsheet so that the calculations and graph updates automatically as I change ambient temperature.

The point of all this is it gives you a decent idea of what the supercharger charge temperatures need to be in order to achieve a give intercooler outlet temperature. What I would like in an ideal world is to have IATs at only 10degC above ambient at WOT. This would keep the ECU well away from seeing temps that require spark retardation on a hot day and therefore reducing power.

As a starting point my spreadsheet tells me that to achieve an outlet temp of 40degC on a 30degC day I need to have supercharger outlet temperatures at about 60degC, or an 83degC reduction from current SC charge temps. Sounds like hard work to me!

It is simple to calculate the energy that needs to be absorbed from the sc charge air to reduce temps to 60degC using the specific heat capacity of air around 1.02KJ/Kg/K - in this case it is 25.88KJ. This does make an assumption that the ambient air is 0% humidity, but I think because of the increase temperature so water carrying ability / reduction in relative humidity this should be relatively insignificant (I hope).

Injecting water cools the charge air by absorbing heat energy in an effort to equalise its temperature with the charge air. It takes 4.2KJ to raise 1kg of water's temperature by 1degC or Kelvin. If the water changes state, in this case from liquid to gas the latent heat of vapourisation also comes in to play. This says that is takes 2260KJ per Kg to convert water at 100degC to steam at 100degC. As my supercharger outlet temperatures would normally be well above 100degC I really think I could really take advantage of this to take the heat out the charge.

I calculated the amount of water in Kg/s required to provide the cooling effect to achieve a range of outlet temperatures, and also compared this flow rate against fuel flow rate to give a required nozzle size in cc/min at the 160psi the WI pumps seem to run at and a water / fuel ratio.

It looks like 40degC outlet temps at 30degC may be a bit out of reach - requiring a water / fuel % of 43% and a nozzle size a fair but bigger than the simple equations on WI supplier sites suggest (1-1.5ccmin/hp). So far I think the max recommended ratio is 25% water to fuel, although I probably need to seek some expert advice on that. That would limit me to an intercooler outlet temp of 48.5degC. 25% also gives me a nozzle cc/min that is comparable to the basic 1-1.5ccmin/hp suggestion. Not quite were I want to be but that would stop timing being pulled. I have also only considered the affect on AIT of the water. The reality is that it also serves to cool the combustion process in the cylinder, and this would give further protection against a 'thermal event' on a hot track day.

I have attached the worksheet I made, feel free to have a play. Hopefully it is laid out in a way that makes sense.

One aspect I know I have overlooked is that the water will also absorb heat from the supercharger body itself. Not sure how much this will affect results but hopefully I have got it accurate enough to spec my nozzle sizes to start experimentation.

If it looks like total ***** please let me know - thermodynamics never was my strong suit!

3rdCarMX5

iTrader: (1)

I haven't checked your math, but it appears that you may be disappointed with just water injected upstream. ​​A lot of the cooling affect will be lost as it passes through the compressor..

​​Water injection on the inlet side of these SCs is supposed to improve the output by 1.5-2psi due to improved rotor sealing, you may be happier with being able to decrease the Rpm (and temps). You could also inject after the intercooler at the same time... But this looks a lot like adding complexity solve for inefficiencies in the overall design ( no offense!).

Have you monitored your intercooler water temps, are those ideal?

Tchaps

Ha ha no offence taken, it kind of is what it is in some respects, and obviously not a particularly well trodden path. I am enjoying the R&D and the discussion, hopefully thread readers are aswell.

Maybe I am being a little close-minded about increasing the number of cores in the intercooler. The more I think about it the more I think I could fabricate the main body, and have end caps machined for not a crazy sum of money. Probably needs a bit of investigation and maybe a quotation from the machine shop. The cores are available so it is just the body that is the issue. It has to be structural, so if I was to make it out of ali I would have to be very careful about how I put it together and what was welded and make it as much as possible from a pre made section.

I have a temperature sensor in the back of the chargecooler. It is a single wire thermistor that is connected to an engine coolant gauge. The scale on the gauge starts at 40degC and I have not yet seen it get on the scale fully. Not really sure on the accuracy of the gauge at that lower end of the spectrum, it isn't really in the temp range it is designed to read. I could do a little better in this respect, but all signs so far haven't shown any unexpected behaviour. Certainly coolant temps do not climb rapidly with prolonged throttle on track. The pump is designed for chargecooler applications so should be adequate, nothing to suggest it isn't currently.

My thinking behind injecting pre compressor is not to let the water pass through it, then cool the air after the supercharger as it is compressed (and heated) at the outlet against the inlet ports - as would be the case with a roots supercharger (no internal compression). The Autorotor is a twin screw with a 1.4 internal compression ratio so it is adding heat to the air charge via compression within the supercharger itself. Injecting the water just upstream of the sc inlet allows the water to absorb the heat as it is being generated inside the supercharger. Take a look at the link below, this article was written by Dennis (from earlier in the thread - he is the Autorotor expert who helped me out with my lubrication system). It talks about injecting fuel or water / meth upstream and its affect on charge temps.

PRESENT-DAY SUPERCHARGING - EXPEDIENCY OR EFFICIENCY? / AJ6 Engineering

I am not so keen on injecting water after the intercooler, in addition or instead of pre sc injection. I don't really have the space and I would have to inject after the AIT sensor, so the ECU would still pull timing reducing the power even if in reality the engine was still safe. I don't think that MAT retard is switchable on MS. I would still be limited by the maximum total water to fuel ratio, so water injected at the ports would have to come out of the amount injected pre sc, where in my mind atleast I see it doing the most good (cooling charge in SC, boosting efficiency, but then still going into the cylinders anyway for the det prevention benefits).

I think you are quite right with what you are saying that with the current intercooler reducing the pulley ratio with pre sc water injection is almost a no brainer - certainly for hot day track use as currently it produces no more top end power on the larger pulleys, just extra heat. Adding water pre sc as I see it currently can only do good - even if it is not as effective as my calculations suggest in the end. It may end up being that even with WI on a hot day I need to use a lower pulley ratio on track, just to keep things nice and safe. At the end of the day it is plenty quick already and it's not like I am entering it in a competitive championship. Reliability and not blowing things up > ultimate power. Still be nice to crack 300hp though :-)