3.0 16V engine mods, etc

[Thom]

I only just completed the install but plan to take the car out tomorrow morning for a short drive as I will be visiting an 88 year old friend who needs a little help to put the engine and gearbox back into his vintage MG.

Yes @Tom, this is a stock turbo mount, though modified to support Garrett GT turbos with "large" turbine housings.

GTX3582R vs GTX3576R :

82_1.jpg (140.6 KiB) Viewed 1263 times

As I said the install was easier than with the '76 as the back plate of the compressor on the '82 is offset frontwards by a few mm and that allows the turbo to move a little more freely around the mount. With the '76 and the latest/largest 1.03 turbine housing used thus far I had to somewhat rub the turbo on and off the mount. Perhaps that small offset can be guessed on the picture below. I was not aware of this minor difference despite having read a lot on these turbos over the years, though perhaps we 951 folks who have always suffered from real estate issues are probably more sensitive to such details than other turbo freaks...

82_2.jpg (139.46 KiB) Viewed 1263 times

All buttoned up and ready to go. No leak, no smoke, no suspicious sound or smell observed after letting the engine warm up at idle to bleed the coolant circuit. Looking forward to the first drive.

82_4.jpg (1.06 MiB) Viewed 1263 times

I have tested different turbine/housing configurations in the past but that was always with 76mm compressor wheels so I'm curious to find out how a larger compressor will affect how the engine runs. On paper at least I do not think that the previous 8V engine would have made the best of such a larger compressor.
IIRC Duke made 767 hp with such a GTX3582R about 10 years ago though that was with very aggressive camshafts. If I can get anywhere close to 600 hp that would be nice for bragging rights but the goal here is first to just test something new and second to at least make the same power as before with less boost.

[Thom]

So... there is lot to say after an initial but relatively short shakedown.

After getting a first feel of how the '82 affected engine response I realised it would need more boost than the '76 in the hope to give anywhere near a similarly dramatic and enjoyable driving experience, so I increased the boost from 1.4 to 1.8 bar/26 psi. The larger compressor makes a big difference in how the car drives as it does take a while to come alive when you floor the throttle but behind its apparent lethargy it flows already more than its little brother did and to get a similar feeling of speed the car must be pushed at an even more ridiculous pace. It is deceptively fast in a way it never was before. Initially I thought I had made a mistake by losing the response of the '76 which made the car an exhilirating drive but the more I drove it the more I started to adjust to the new identity of the car and I now quite like it, despite the clutch unable to handle the new level of performance - the log file below is worth a look, going full load in 3rd and 4th and holding back in 5th as the clutch started to slip and smell like crazy...

Even after maxing out the values in the fuel map the mixture remains too lean at an AFR of ~13:1. The injector duty cycle with the '82 is peaking at ~100% which is about 10% higher than with the '76. I did not have much time to drive the car extensively to retune it "completely" but I will have to at least increase the Master Fuel (base pulse width) and redo the fuel map. That should be easy to recalculate, at least to get a decent starting point, once I find back the fuel equation the Vi-PEC software uses.

log1.jpg (131.79 KiB) Viewed 1200 times

log2.jpg (133.83 KiB) Viewed 1200 times

log3.jpg (135.09 KiB) Viewed 1200 times

Ambient temp was fairly high at 31°C, and the compressor seems to be pushing slightly out of its peak efficiency range with an inlet temp of +14°C which I think is still very reasonable. EGT is way below what it was with the '76. Perhaps the most interesting observation made when recording this log is that even with such a lean mixture and such a high boost level there was absolutely no audible knock listening to the engine through the J&S. I am pretty sure that back pressure is at an all time low and might even be lower than boost pressure considering the low EGT, lack of detonation and the massive performance altogether. It all seems a little surreal.

As I mentioned in a previous post the goal was to get more flow with less boost which I think the '82 provides in spades, and a side result is that cruising under light and medium loads gives more performance thanks to the '82's larger compressor wheel that doesn't need to spin as quickly as the '76's. Clearly the '82 is a large beast that takes a while to wake up but even under transient condition it already offers more performance. Even manoeuvring the car in 1st or reverse the engine is a little less jerky, suggesting there is more air flowing under any sort of load whilst the '76 needed a bit of "cranking" with the throttle... FLOW is the $hit

Now what sort of streetable clutch can I find to regain traction...

[Tom]

Wow, you work quick. 26psi and 7k rpms.... Car must be a rocket ship! I think we need a video.

What clutch do you have now and how much torque do you think the motor is making?

[Thom]

I'm still on a KEP2 PP with apparently a stock 911 3.2 disc which had been sold to me as a 930 disc...

ETA : the clutch disc is actually a later style, 4 spring 930 disc, which replaced an early style, 6 spring 930 clutch disc.

[Tom]

You might call KEP and see what they suggest. I'd be worried about your firewall (unless reinforced) if you added any more clamping force -- so you might need a disk with more friction. I wonder if you could get by with a resurfaced flywheel and new 951 cup disk? Yours has lots of, um, 'testing' so might not have the grip it once did -- and the 3.2 disk is kind of a wildcard...

If that's not enough and you need a disk with a higher friction rate, consider South Bend clutch. They have a unique 'feramic' friction material that adds considerable holding power while still retaining 'near' stock engagement feel. I ran one for a while. On a smoothness scale, if stock is 10 and Spec is negative three million, the South Bend disk is about an 8 I'd say. South Bend can also modify a Sach PP to add as much clamping force as you want.

[Thom]

Thanks Tom.
If anything I'd like to keep the KEP2 PP and the TTV flywheel which where balanced together. Will South Bend supply a disc alone? On their website they seem to supply full clutch kits only.

[Tom]

Thanks Tom.
If anything I'd like to keep the KEP2 PP and the TTV flywheel which where balanced together. Will South Bend supply a disc alone? On their website they seem to supply full kits only.

I haven't bought anything from them in a long time, but when I did, they were willing to make just about whatever you needed. They were more of a service than a product supplier. Their website isn't great, but at least at the time everything was pretty much made to order... I'd call them. Hopefully, they haven't lost their way. KEP can also do disks, so also worth a call if you are keeping your PP, though I don't think they have the range of friction-material options SB does. Could be wrong about that.

[Bergerac]

Time for some bigger injectors! What are you running now?

[Thom]

Bosch EV14 1200cc

[TC72]

Even after maxing out the values in the fuel map the mixture remains too lean at an AFR of ~13:1. The injector duty cycle with the '82 is peaking at ~100% which is about 10% higher than with the '76. I did not have much time to drive the car extensively to retune it "completely" but I will have to at least increase the Master Fuel (base pulse width) and redo the fuel map. That should be easy to recalculate, at least to get a decent starting point, once I find back the fuel equation the Vi-PEC software uses.

You shouldn't be hitting 100% duty of 1400cc injectors unless you are approaching 800hp(assuming gasoline and not E85). 283kpa absolute with a ~220hp base engine should put you around 630hp - not sure if you have upgraded cams, but this estimate should be +-10%. Do you log fuel pressure? What pump are you running?

With the link/ Vipec fuel model, the values in the base fuel table are percentage of the master fuel, but your log shows actual duty cycle, so increasing the Master Fuel will not get you additional fuel. The maximum pulse width confirms this - 7000 rpm is 116rps, two revolutions per cycle is 17.2ms available to inject.

The link/ vipec model is just percentage of the Master Fuel setting. They also have the "Modelled" fuel mode, which is essentially VE with charge temp approximation, but this does not have a Master Fuel setting. If you increase your master fuel setting, you should be able to scale the base fuel table by the same percentage(if you go from 4ms to 5ms, decrease the base table by 25%) then clean up the non-linear areas where pulse width is very low or high.

Looks like a pretty positive result- new turbine wheels perform so well in terms of back pressure. It really makes high boost on pump fuel more practical, and engines really love it.