![]() |
|
|
|
|
![]() |
|
LinkBack | Thread Tools |
Rating: ![]() |
Registered User
Join Date: Apr 2011
Posts: 235
|
Ni-resist rings for 100mm, how do they work?
__________________
1988 930: 3.5L 100mm LN slipins, LN FFA rods, Supertec studs, GT3582R with T4 1.07A/R twin scroll housing, RarlyL8 equal length divided headers, dual Tial MVS 38mm WG's, Carrera 3.2 heads with 993 big valves. 964 Cams, full bay IC, KEP st1 clutch. PE3 sequential ECU with 750cc EV14 injectors, CoP twin plugs, J&S Safegaurd, Aquamist HSF3 water inj. Wavetrac LSD. Last edited by b930; 06-04-2017 at 02:20 AM.. |
||
![]() |
|
Chain fence eating turbo
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,726
|
Watching. I'm in the same boat. Some feel you can seal by a lot of preload torque and freshly machined surfaces, others feel you need flame rings or use a gasket system.
I too had been running 1.4 and spiked to 1.6. |
||
![]() |
|
Registered User
Join Date: Jan 2006
Posts: 174
|
There is certainly a more modern approach to this issue available. The old Ni resist rings are from the 80's.
Look at performancedevelopments.com – Engine Design. under their tech corner. They have what you are looking for. |
||
![]() |
|
Registered User
Join Date: Jan 2006
Posts: 174
|
Also, take a look at Garlock ring. They are the gas filled type you mention. Not sure who sells them.
|
||
![]() |
|
Registered User
Join Date: Apr 2011
Posts: 235
|
Interesting insights from performance developments site:
Some of the early 911 Turbo engines used a stainless fire ring between the cylinder and the head. This ring was not very good at sealing the two parts as it was primarily there to stop the flame travel from burning its way out. This primitive system also did not stop the head from banging against the cylinder top. Our system uses some interference between the sealing gaskets and the head and cylinder. This interference compensates for any head lift due to pressure and temperature.
__________________
1988 930: 3.5L 100mm LN slipins, LN FFA rods, Supertec studs, GT3582R with T4 1.07A/R twin scroll housing, RarlyL8 equal length divided headers, dual Tial MVS 38mm WG's, Carrera 3.2 heads with 993 big valves. 964 Cams, full bay IC, KEP st1 clutch. PE3 sequential ECU with 750cc EV14 injectors, CoP twin plugs, J&S Safegaurd, Aquamist HSF3 water inj. Wavetrac LSD. |
||
![]() |
|
Rocket Scientist
Join Date: Jun 2008
Location: Detroit
Posts: 908
|
What's wrong with the 993 CE rings?
|
||
![]() |
|
Registered User
Join Date: Feb 2004
Location: UK
Posts: 2,231
|
Any idea of the peak cylinder pressures?
|
||
![]() |
|
Registered User
Join Date: Jan 2006
Posts: 174
|
Without knowing the spec of the engine that would be hard to estimate. If his engine is making torque in the region of 450+ Ft/lbs with a comp ratio of 9 @ 22psi MAP, he could be seeing upwards of 120b without a leak.
So many factors could lower this, cams, head flow etc. Best guess would be somewhere between 90 - 120 maybe. That's a lot to expect the heads to seal without any sort of gasket. With a leak who knows. |
||
![]() |
|
Registered User
Join Date: Apr 2011
Posts: 235
|
Tippy, yours leaking also? I think I am going to send my parts to performance developments. I quite like their solution.
__________________
1988 930: 3.5L 100mm LN slipins, LN FFA rods, Supertec studs, GT3582R with T4 1.07A/R twin scroll housing, RarlyL8 equal length divided headers, dual Tial MVS 38mm WG's, Carrera 3.2 heads with 993 big valves. 964 Cams, full bay IC, KEP st1 clutch. PE3 sequential ECU with 750cc EV14 injectors, CoP twin plugs, J&S Safegaurd, Aquamist HSF3 water inj. Wavetrac LSD. |
||
![]() |
|
Chain fence eating turbo
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,726
|
Quote:
Looked like an oil leak going up instead of down. Mine was from using 3.2 head torque instead of 993TT, which are what my studs are. |
||
![]() |
|
Rocket Scientist
Join Date: Jun 2008
Location: Detroit
Posts: 908
|
For design purposes such as this we can just use a couple of estimation methods.
First, the BMEP in kPa is (6.28 x 2 x torque)/Displacement, with torque in Nm and displacement in liters. For a turbo motor making 500 HP at 6800 RPM, the torque is 386 ft lbs or 523 Nm. Plugging that in the BMEP formula for a 3.4L engine: (6.28 x 2 x 523)/3.4 = 1932 kPa or 280 psi. BMEP is the constant pressure that, if applied during the power stroke would produce the brake torque. We are not really interested in the BMEP though. We are interested in the peak firing pressure. That is what lifts the heads and gives the head studs a workout. That takes some figuring with the Weibe function, or some cylinder pressure data. I have done it for a 3.2 Turbo motor, and it came out to be about 4 times the BMEP, so I would hazard a guesstimate of about 280 x 4.0= 1120 psi for the peak firing pressure. Last edited by Speedy Squirrel; 06-06-2017 at 10:25 AM.. |
||
![]() |
|
Registered User
Join Date: Feb 2004
Location: UK
Posts: 2,231
|
I am not sure how the interlocking Ni-Resist Ring still seals if the heads lift.
I can't find any cross sections of the grooves and how the seals locate. I understand the Wills Ring - originally known as a Coopers Rings and have used the MOT style for many years. If the head lifts sufficiently for this ring to lose its preload then it will still leak so the extent of the lift is quite significant to using these parts. If we assume 120 bar then the force trying to lift the head will be around 5500lbs per stud. I would estimate the preload in a stud torqued to 24lbsft to be around 5000lbs so if heads were torque to this level then leaks seem to be inevitable. If we torque the studs to 36lbsft then the preload should be around 7000lbs so if the heads still leak then the preload needs to be greater. or a better seal is required. Using an MOT type ring to gain a pressure seal at around 5000psi would ideally need the ring to be seated with a preload of about 300lbs per inch of circumference. I think using a 1.59mm dia ring then if you can preload at the higher level these will still be enough preload to seal but if the head lifts completely they will still leak. I have always wondered why the CE rings are used in the wrong direction as normally the 'C' would be open to pressure and this would help with the sealing even if the heads did lift slightly. I guess that they must collect carbon deposits and cause pre-ignition. |
||
![]() |
|
![]() |
Chain fence eating turbo
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,726
|
Awesome info, again, Chris!
Why not use dead soft copper rings? Easy to punch out and seal really well. |
||
![]() |
|
Registered User
Join Date: Feb 2004
Location: UK
Posts: 2,231
|
I think that the Wills Rings is a better bet as the internal gas pressure helps with maintaining the seal.
I don't think that the Nitrogen helps with expansion but it helps the ring to maintain its resilience when hot and compensates for the reduction in Young's Modulus. We use a 321 Stainless Steel Ring rather than Nimonic and they will easily withstand the temperatures. I think using a copper ring would still need a high contact force to maintain a seal and it seems as if the reduction in preload that occurs with the high peak cylinder pressure is the problem but I am sure that there are others with more knowledge in this area. The Wills Ring will tolerate this reduction in load as its has more capacity to deform elastically which must help maintain the seal. There's lots of information here: https://tss-static.com/remotemedia/media/globalformastercontent/downloadsautomaticlycreatedbyscript/catalogs/wills_rings_gb_en.pdf Last edited by chris_seven; 06-07-2017 at 10:55 AM.. |
||
![]() |
|
Chain fence eating turbo
Join Date: Dec 2008
Location: Austin, TX
Posts: 8,726
|
^^^ Good stuff
|
||
![]() |
|
Registered User
Join Date: Jan 2006
Posts: 174
|
Regardless of what type of sealing ring type is used, you must have control of the assembly procedures. Often these are taken for granted.
Once the parts are mated together and zero gap, you cannot change the thickness of the parts unless some collapse under the compression pressure. This should not be happening and the forces should be going into the stud stretch, to act as tensile force clamping and holding the parts together resisting all of the pressure forces trying to push them apart. You have to know the stretch of the stud being used at whatever torque you use. You should be using an angle method to remove any friction changing the clamping load. Grease type and or oil type can change this as well. If the stud does not stretch, then either the stud starts to pull out of the block or the head collapses under the head washer. Resulting in different clamping loads on each stud. Whatever stud you are using, you need to know the stretch from the supplier or at least test before installing. The cylinder pressures in these engines are very high and usually unknown to most. When there are leaks and head lifting, its always blamed upon the stud, when in fact the stud may not be made for this application and the basic assembly procedures are not applied. |
||
![]() |
|
Registered User
Join Date: Apr 2011
Posts: 235
|
So after much research I deduced to go with Performance Development head sealing system. Generally building proper race motors with years of experience, all I can say is these guys know their stuff. Not cheap but you pay for what you get. Their solution is like nothing else on the market.
They measure torque / stretch of the studs and machine both cylinders and heads. The gaskets are EDM machined to very fine tolerances. ![]() ![]()
__________________
1988 930: 3.5L 100mm LN slipins, LN FFA rods, Supertec studs, GT3582R with T4 1.07A/R twin scroll housing, RarlyL8 equal length divided headers, dual Tial MVS 38mm WG's, Carrera 3.2 heads with 993 big valves. 964 Cams, full bay IC, KEP st1 clutch. PE3 sequential ECU with 750cc EV14 injectors, CoP twin plugs, J&S Safegaurd, Aquamist HSF3 water inj. Wavetrac LSD. Last edited by b930; 08-11-2017 at 05:44 AM.. |
||
![]() |
|
Registered User
Join Date: Mar 2004
Location: So. Ca.
Posts: 453
|
|||
![]() |
|
Registered User
Join Date: Feb 2008
Location: Marietta GA
Posts: 2,559
|
How much is "not cheap" in dollars?
|
||
![]() |
|
Registered User
Join Date: Feb 2004
Location: UK
Posts: 2,231
|
Quote:
There is a 997 part described as a Garlock Ring and another hint at what sounds like an MLS system. I am very interested to learn more |
||
![]() |
|
![]() |
Thread Tools | |
Rate This Thread | |
|