Regulations usually only concern safety. While that chassis may be perfectly fine as-is on the street, adding a stressed panel across the bottom would significantly increase its torsional stiffness. in fact, TUEV noted that "...they complain that the torsional stiffness in the rear is very low..". There you are. QED.
VW W12 to Porsche G86.20 transaxle
- Thread starter CTCH
- Start date
Thanks for the reply. Can you explain what you mean by a stressed panel? Do you mean a sandwich panel? Please let me know.
A stressed panel would be a sheet of aluminum riveted securely to the bottom chassis tubes. Something like this:
The panel resists the chassis twist.
Absolutely, that makes sense. I am considering using a 5mm CFRP sheet for that:
www.carbon-composite.com
Unsure if this is common?
CFK Carbon Platte 1300mm x 900mm x 5.0mm
✅ Hochfeste CFK Carbonplatten ✅ Carbon Sandwich Platten ✅ CNC Bearbeitung ✅ gefräste Bauteile ✅ Sonderanfertigungen ✅ Hochmodul ✅ Hybrid Platten ✅ Kohlefaser ✅ Carbonfaser ✅ Glasfaser GFK ✅ Hochtemperatur
www.carbon-composite.com
Unsure if this is common?
Using 2024-T3 x 0.062" thick aluminum will save you $$$$$$$$$$$$.Absolutely, that makes sense. I am considering using a 5mm CFRP sheet for that:
CFK Carbon Platte 1300mm x 900mm x 5.0mm
✅ Hochfeste CFK Carbonplatten ✅ Carbon Sandwich Platten ✅ CNC Bearbeitung ✅ gefräste Bauteile ✅ Sonderanfertigungen ✅ Hochmodul ✅ Hybrid Platten ✅ Kohlefaser ✅ Carbonfaser ✅ Glasfaser GFK ✅ Hochtemperatur
Unsure if this is common?
View attachment 138042
Kennedy Engineered Products, Inc. said: "We adapt the Small Block Ford, Ford Coyote, Ford Ecoboost 2.0/2.3, Small Block Chevrolet, Chevy LS, LT4, LT5, Chevy Ecotec 2.2/2.4, Honda K20/K24, and Toyota Supra 2JZ."
Checked the torsional stiffness of that frame today. The front left damper attachment point is pulled up with 1 ton here, whereas the other 3x damper attachment points are not allowed to move. The front left point of the frame deforms almost 70mm according to that model.
Stresses exceed 300MPa in a lot of the corners. Likely staying permanently deformed after such a load. Would love to compare a GT40 frame to that here. Does anyone have a CAD model of a GT40/replica frame?
Stresses exceed 300MPa in a lot of the corners. Likely staying permanently deformed after such a load. Would love to compare a GT40 frame to that here. Does anyone have a CAD model of a GT40/replica frame?
It looks like the front bay is intended to have panels on the non-diagonalized areas.
The vertical tube at the front of the rocker areas really needs more transverse rigidity.
The vertical tube at the front of the rocker areas really needs more transverse rigidity.
Thanks for the comment, tried some panels - makes a significant difference.
Added a bottom "stressed panel" (1mm steel or 3mm aluminium)
Added side panels then additionally
Added panels and additional members on the front and at the rear
Outcome with respect to baseline - almost doubled the torsional rigidity.
Does somebody have a GT40 replica frame? I would like to run a comparison.
Added a bottom "stressed panel" (1mm steel or 3mm aluminium)
Added side panels then additionally
Added panels and additional members on the front and at the rear
Outcome with respect to baseline - almost doubled the torsional rigidity.
Does somebody have a GT40 replica frame? I would like to run a comparison.
Frame with all "stressed panels" and additional members comes in at 160kg in steel. Not super light weight - how does that compare to the GT40 replicas?
Attachments
Now buy yourself a pneumatic/hydraulic rivet gun and about 1,000 AVEX or CherryMAX blind rivets,,, and a bunch of #30 twist drills.
I think that 1.5mm 2024-T3 aluminum would be a good choice for those stressed panels.
Some more numbers on torsional stiffness: The base frame has about 2300 Nm/degree and the fully paneled one has almost double with 4400 Nm/degree.
Original Countach had 2600 Nm/degree (also had no panels, only tubes)
McLaren F1 had 13500 Nm/degree
Porsche 911 Turbo (2000) 13500 Nm/degree
Lamborghini Murcielago 20000 Nm/degree
sources:
Found this statement on the web that resonanted with me "Once you get over 5000 then you can start to tune around the soft chassis (for a road car) and once its over 15000 Nm/deg frankly you are just playing specmanship or have other (possibly useful)motives."
So 4400 is roughly 70% stiffer than the original Countach, but still on the soft chassis side of things.
Original Countach had 2600 Nm/degree (also had no panels, only tubes)
McLaren F1 had 13500 Nm/degree
Porsche 911 Turbo (2000) 13500 Nm/degree
Lamborghini Murcielago 20000 Nm/degree
sources:
Found this statement on the web that resonanted with me "Once you get over 5000 then you can start to tune around the soft chassis (for a road car) and once its over 15000 Nm/deg frankly you are just playing specmanship or have other (possibly useful)motives."
So 4400 is roughly 70% stiffer than the original Countach, but still on the soft chassis side of things.
Ole Nylende
Supporter
I think i have some numbers from Malmberg also. Might be worth to have a look at the speedlab corvette build also, i seem to recall that they did some testing on that chassis.
Add a roll cage and the torsional stiffness goes way up.Some more numbers on torsional stiffness: The base frame has about 2300 Nm/degree and the fully paneled one has almost double with 4400 Nm/degree.
Original Countach had 2600 Nm/degree (also had no panels, only tubes)
McLaren F1 had 13500 Nm/degree
Porsche 911 Turbo (2000) 13500 Nm/degree
Lamborghini Murcielago 20000 Nm/degree
sources:
View attachment 138139
Found this statement on the web that resonanted with me "Once you get over 5000 then you can start to tune around the soft chassis (for a road car) and once its over 15000 Nm/deg frankly you are just playing specmanship or have other (possibly useful)motives."
So 4400 is roughly 70% stiffer than the original Countach, but still on the soft chassis side of things.
Got the mounting plate in between the W12 and G86.20 figured out. 3x of the bolts match, but the adapter plate is still needed. I kept the thickness low - its 10mm. The green plate is the adapter plate. The blue geometry is the W12 flange. Also got a supplier that can make the flywheel, etc figured out.
Ian Anderson
Lifetime Supporter
Reach out to Chris on here he is Flatchat, He is based in Australia and has made a few adaptors for different engine / transmission combinations
Ian
Ian
That is exactly the gentleman I am working with!
Removing the starter is really a nightmare on these engines - it requires taking the headers off! I tried without, but no luck. As such, my idea is to mount the starter wheel on the harmonic balancer and relocate the starter.
Double checked everything and asked Chris to machine the flywheel and adapter plate for me. The adapter plate (in green) mounts to the G86.20 with countersunk bolts. The G86.20 with the mounted adapter plate then mounts to the W12. I wanted the mounting plate thickness to be 0 and mount the gearbox directly on the engine, but only 3x of the bolts in the bolt pattern of the W12 match those of the G86 and the overall solution to make this work would have been ugly. I can live with 10mm - the entire reason why I love this W12 is its compactness. The camshaft chains are "in the back of the block overhanging over the flywheel mounting flange" and a thick adapter would negate all of these advantages. Flywheel is 47mm in height - reasonable enough. Chris suggested to use a 8" Uniclutch - which does not need a hard running face on the flywheel. This suits quite well because of two reasons: it is a dual clutch design and therefore matches the torque required with the small diameter of the G86.20 bell housing. Second, it can mount on an aluminium flywheel - and since the thickness of the flywheel is 47mm, I really do not want to have this made in steel. Love it, now let us hope that the computer work and the real product are not miles apart from each other!
Earl Balzen
Supporter
