Project Turbo Killer

[sonett]

Maybe you guys over in the US have already seen my thread on UKSaabs, but i thought i should post something more permanent for reference, the full thread is here http://www.uksaabs.co.uk/UKS/viewtopic.php?f=35&t=17159

The background of this project was to get my 1983 99GL above the horsepower of a standard 99 Turbo, without using forced induction and to also have a fast road car that could be easily driven with a wide power band. As we all know the 99 turbo puts out 145bhp, obviously the torque figure could not be reached.

Throughout the project a rebuilt late 900i H block was used, the type with the starter motor under the inlet manifold, the oil pump was changed for the later 3 lobe type. Bore is standard, along with crank and pistons, new piston rings, bearings, timing chain and an adjustable top sprocket were fitted. The exhaust system is a 4-2-1 manifold from MSS coupled to a 900 front and rear silencer from Jetex, the tail pipe is Ansa. I did think about ditching the rear silencer but this will be a road car so i needed to think about some comfort and noise. I do have some concerns about the MSS manifold, more about this later.

The first state of tune was a standard head with no porting and standard valves, although compression was increased to around 10:1, fuelling was Bosch K-Jetronic and the cam was a reprofiled Piper cam with a duration of 272 with 11.42mm lift on the dyno this was giving 127bhp @ 5800rpm with 124 lbft @ 4500rpm. I used the car for some time like this and it was good to drive, not exactly blistering performance but OK.

The second state of tune was using a Puma Race Engines head, gas flowed and 44mm inlet valves (standard inlets are 42mm) standard injection exhaust valves, camshaft is a Catcam 0022, 256*@ 1mm, 284*@0.1mm, total lift 12.2mm. Fuelling was again Bosch K-jet. This was not as good to drive as the first state of tune, nothing much below 2000rpm then it really picked up, not easy to drive in traffic. On the same dyno it was giving out 135bhp @ 5900rpm and 133 lbft @ 5000rpm. This was the point when power was becoming restricted, instead of power peaking and tailing off it just flat lined from around 5000 rpm all the way to over 6000rpm which is where we cut the run anyway, this is usually the result of an exhaust manifold restriction, but not in this case. I decided to fit the K-jet fuel head/sensor plate from a 99 Turbo to see if it helped. It certainly helped low down and the induction noise was higher, it also pinked slightly at high load, getting a nice idle was also tricky, i fitted an adjustable warm up regulator and took it back to the rollers to see if things had improved. Power was up slightly to 140bhp @ 5900rpm torque was much the same, but it still flat lined at peak power. We then fitted a vacuum gauge to the inlet manifold to see what was happening at WOT, it was still pulling 3-4psi vacuum, so the engine was asking for more air, but the K-jet couldn't deliver it. This actually seems to tie in with the Saab rally cars, the early K-jet rally cars were all around 130-140bhp, then the next step was over to twin carbs, i can now see why. Dave Baker (Puma Race Engines) ran some figures and came up with the following;

In an ideal world a perfect induction system would have almost zero manifold depression at WOT. Any restriction there means a reduction in the amount of airflow into the engine. When American carb manufacturers started flow testing their carbs to give a CFM (cubic feet per minute) rating they settled on a flow bench test pressure of 1.5" of mercury. The reason was that it had been found that if manifold depression exceeded that amount power started to suffer. In other words their flow figures were saying you can use this carb to supply that number of CFM of airflow and it won't cause a manifold depression greater than 1.5" of mercury.

Atmospheric pressure (14.7 psi) is about 30" of mercury so 1" of mercury is about 0.5 psi. 1.5" of mercury is therefore about 0.75 psi and that's as much as we ideally want to see at WOT. 0.2 to 0.3 bar is a huge manifold depression. 1 bar is 14.5 psi so we're looking at 2.9 to 4.35 psi. That's reducing the net atmospheric pressure feeding the engine from 14.7 psi down to 11.8 to 10.35 psi.

We can calculate the effect of this on power. Flow, and therefore power, is proportional to the square root of pressure drop. If we take 14 psi as being representative of a good non restrictive manifold i.e. about 0.75 psi below atmospheric, then we're looking at a pressure ratio of between 11.8/14 to 10.35/14 = 0.84 to 0.74.

Taking the square root of those we're looking at a flow and power ratio of 0.92 to 0.86.

In other words we're losing between 8% and 14% of the potential engine power. That means the 140 bhp should really be somewhere between 152 and 163 bhp. The average of those is pretty much exactly what I originally calculated the engine should produce with a good induction and exhaust system.

What you really now need is a flow bench so you can test each part of the induction system and find where the restrictions lie. Or just stick the TBs on.

Dave Baker
Puma Race Engines

I was pleased and dissapointed all at the same time, pleased to know there was nothing wrong with the engine and dissapointed with the K-jet, i almost reached my goal of 145bhp, although with a car that was not easy to drive everyday.

State of tune number three and the final state (I think) same Puma Race Engines head etc, but this time with a Misab inlet manifold and 45mm Jenvey throttle bodies, i started off with megasquirt as the ECU, but i had many problems with this, more later, so i went for a DTA S40, once i finished the wiring it started using a map for a Ford Zetec, it ran but not perfect but enough to get it back to the dyno for a remap. The mapping was done by Wayne Schofield (Chipwizards). Because of the low vacuum at idle it was best to use alpha-n, which is basically using rpm and tps for the load sites (tps=throttle position) I was actually with Wayne the full day, his attention detail is excellent and the result proves the fact. The inductive pick up on Waynes dyno was damaged so a torque graph couldn't be plotted, but peak power is 166.6@6400, so eventually we smashed the 145bhp target on a relatively mild camshaft. The car is absolutely brilliant to drive and very flexible, i can put it in 4th gear at 1500rpm and it will still pick up revs, when at cruise the throttles are almost closed so it just sounds like a standard car with no noise, press a bit more on the accelerator and the noise is great. We could have squeezed out a bit more power and maybe broke the 170 mark, but Wayne dialled it back a little to be on the safe side, after all it is a road car. I kept the rpm limit at 6500rpm because the valve train is still using standard parts, the actual power curve was still climbing at this point!

Whilst at Wayne's dyno i wanted to play about with the induction length, but time was not on our side, so a few days later i took it to another local dyno (Dastek) to have a play with the induction length. The throttle bodies had 40mm trumpets and an ITG filter bar fitted when it was mapped, the first run on the dyno i removed the filter bar and this only effected power by a few hp, then i changed the trumpets to 90mm, again a few hp at peak was sacrificed but it was worth it because the torque was coming in 500rpm earlier. I just need to fabricate a back plate for the 90mm trumpets so i can get the filter bar back on

Back to the MSS 4-2-1 manifold, this is a good manifold and as you can see it works well, but i can't help but think about the primary and secondary pipe diameters, they are both the same, every 4-2-1 manifold i have seen has smaller primary pipe diameters than secondary, i have a spare MSS manifold that i am hoping to change the primary pipes for a smaller diameter and see what happens to the power/torque figures.

[Geoff]

Wow, that's pretty impressive! I started reading the thread on UK SAABs but its really long (and BIRDIE is tough to read sometimes! ) I'll look at the pictures tonight.

Did you measure the intake runner length? Did you do any calculations on runner length for optimum power or figure out which harmonic that length of runner is taking advantage of?

I'm not sure why Jack used the same pipe diameters on the primary and secondaries. He did develop the headers using his own dyno ("RR") so I'd think there would be a reason for it. I've talked with him a few times about headers. He also made a race header that used larger diameter pipes and the secondaries were shorter (not sure about the primary length). I bought a 16v header from him and I asked him what I should do when I put it on a race engine. He told me to start by shortening the secondaries "to about here" (about 1 foot or so). He could probably tell you why its designed the way it is...

[sonett]

No calculations were carried out for optimum runner length, have a look at the bottom of this page http://www.pumaracing.co.uk/gentune.htm once again Dave was correct, i fitted longer trumpets and the torque came in 500rpm earlier than with the 40mm trumpets. Even if calculations were to be done there is not much you could do anyway, apart from alter the trumpet lengths because everthing else is more or less fixed such as the inlet manifold and bodies.
However i did measure the inlet tract and the overall length is 415mm or just over 16'', which is a pretty good inlet length anyway.
As you can see from runs 5 and 6 the torque has been moved down with the 90mm trumpets, also note the power curve is still rising!


I'll drop Jack a line and ask him about the pipe sizes, it's just puzzling that it does contradict the reason for different diameter pipe sections.

[Luke]

MSS race header vs. street header:

[sonett]

Thanks for the pics Luke.
That race header just doesn't look right, it would be good to see a back to back test on both of those manifolds on the same engine and dyno.
I've emailed Jack to see if he has any information on the headers such as dyno runs etc.

[SwedeSport]

I have actually had both on my 8v project. I got the street header with O2 sensor bung second hand from James Fox a few years back. Sand blasted and re-coated the thing with good header paint and wrapped it with exhaust wrap. (big mistake on a street car). The secondaries down rusted out in a matter of about 2 years. Jack was kind enough to supply me with a new set pipes to replace my damaged stuff. But this time I opted for the race header secondaries. Put it all together and put it back in the car. This time with no header wrap. It seems to be slightly better in the low rpm range as well as in the mid range but the car is rather flat up top. The configuration of the remaining exhaust has changed over the years. It has a stock 85 16v turbo tailpipe, a cherry bomb muffler (flowmaster copy) 2 1/4 center pipe, and the cat was replaced with a straight thru bullet muffler. I tried to run with just a straight pipe but it was too loud. It would pop and backfire on decel and it would set off car alarms in the vicinity. Its definitely better than the stock system, but Im sure could be improved.

[Jordan]

A point I'd make is that the race header is really a 'tuned' header that is part of a system (cam, intake etc) and without the other parts of the system it isn't going to have the same effect. When I had the race header on my stock engine I noticed better throttle response, but that was about it. I wasn't expecting much because I hadn't fitted the rest of the system yet. Not to say maybe the design couldn't be improved at all, but I don't think you can do piecemeal analysis on just one component.

FYI, Jack rang out ~ 200hp out of his 8v carb'd motor in the end, but he said it was extremely finicky, and the power band was very narrow...

[sonett]

FYI, Jack rang out ~ 200hp out of his 8v carb'd motor in the end, but he said it was extremely finicky, and the power band was very narrow...

That is exactly what i had to avoid, a car with a very narrow power band is bad enough to use in competition, let alone as an everyday car.

[sonett]

After discussions with Dave Baker (www.pumaracing.co.uk) he advised i should fit smaller primaries, he knows a lot more about this than me, so it would have been foolish to ignore his advice, according to Dave 44mm primaries would come into play at around 220bhp. I swapped the 44mm primaries for 39mm, fastened it all up and took it to the same rollers.
The first run on the rollers had the 40mm inlet trumpets fitted, peak power was down by a few bhp, no big deal, but then we overlaid the graph from a previous run with the old exhaust manifold and same size trumpets, low down torque up to around 3500rpm had jumped up, at 2500rpm it has increased by around 7lb/ft.

[sonett]

OK, so much for the short trumpets, Andrew was in no rush to go home and he seemed as interested as me to see what the 90mm trumpets would produce, 5 minutes later the longer trumpets fitted and we're ready for another run.
Comparing this run with a previous run, with the old manifold and 90mm trumpets, peak power was up by a tad over 6bhp, but this time it kept rising, i told Andrew not to go over 6500rpm, so 166bhp was it's limit at 6300rpm, maybe we could have reached 170, but in this case i am happy with what i've got, a very respectable fast road car :).
Torque also increased, almost throughout the entire rev range and peak torque was up by 6lb/ft.
OK, so much for the short trumpets, Andrew was in no rush to go home and he seemed as interested as me to see what the 90mm trumpets would produce, 5 minutes later the longer trumpets fitted and we're ready for another run.
Comparing this run with a previous run, with the old manifold and 90mm trumpets, peak power was up by a tad over 6bhp, but this time it kept rising, i told Andrew not to go over 6500rpm, so 166bhp was it's limit at 6300rpm, maybe we could have reached 170, but in this case i am happy with what i've got, a very respectable fast road car :).
Torque also increased, almost throughout the entire rev range and peak torque was up by 6lb/ft.

[sonett]

To conclude, the smaller primary pipes have made a difference, but the difference was improved further by using the 90mm trumpets. Here is a final graph with the new exhaust manifold comparing the 40mm and 90mm trumpets.
I have since learned about two manifolds available from Sweden, similar in design to what i already have, but with larger pipes, 42mm primary/48mm secondary and 60mm exit pipe, this one is intended for 2150cc rallycross, the other even larger with 44mm primary/51mm secondary and 64mm exit pipe, intended for 2400cc rallycross, but unless you go with a big valve 16v head these will be of no benefit.
The MSS manifold is quite decent and is much better than the standard cast iron unit, but making the primaries smaller has yielded worthwhile results, they have lifted the low down torque a good chunk with the smaller trumpets, but i have gained 6bhp and 6lb/ft torque at the top end with the longer ones.

[Rallyho]

That sounds very close to what I am looking to do with a 16v head.

I have a very similar looking MSS header for the 16v. A set of ITB that Luke crafted. Some forged flat top pistons that I hope will fit with minimum machining. (valve pockets) and a plan to run some sort of programable fuel infection.

I'm not familiar with the FI you used. Can you explain some more about it and what your difficulties with MEgasquirt were?

thanks...awesome results with real math behind them...love it!

Mike

[sonett]

It can be quite frustrating reading some forums when you are looking for information, you often read things like 'i fitted a bigger exhaust and the car felt quicker'.
Here is a link to the ECU i am using www.dtafast.co.uk it's very popular over here and lots of rally and race cars use them, the majority of mappers are very familiar with them and they also supplier a tuning box, this consists of two knobs, ignition and fuel, so you can alter these variables very easily on the dyno at a sustained RPM and see the effects on power/torque.

I bought my MS from a professional supplier here in the UK, the spark driver burnt out, the codes were not correct, when i finally got it running the ignition timing was wandering about at random, i removed it, fitted the DTA and it fired up first time on a map i chose from the menu in the ECU.

My next Saab engine project, after i get the ultra big valve head B engine running, will be a 2.1 16v n/a engine, this will be on throttle bodies and taken out to 2500cc, the head will be done by Puma Race Engines, with the biggest inlet valves we can fit

[sonett]

Dave Baker has added two useful sections to his website;

www.pumaracing.co.uk/manifold.htm

www.pumaracing.co.uk/exhaust.htm

[SwedeSport]

I actually got some real dyno numbers for my 2.1. 8 valve . I didn't actually use the cylinder head I planned because I need to have a cam made to use it. I took a 2.1 bottom end out of 150k parts car, dropped on a mildly port/polished head with dual springs and a cam. I managed to pull 109.10 HP at 5500 rpm, and 111 ft-lb at 4500 rpm. This is at the wheels. Not too disappointed at that, it was afterall a bunch of used parts that I threw together.