Dan@SharkWerks

Hi guys,
I cross-posted this from the GT3 section but this issue applies to 996 Turbo models as well so I thought it would be a good idea to post here.

Some of you may be aware of coolant pipe issues on GT1-block equipped models (GT3, GT2, Turbo), where coolant pipes come apart while driving and the rapid loss of engine coolant can cause spins/crashes at race tracks when slippery coolant sprays all over the rear tires.

The problem exists on the GT1 motor because there are a couple coolant pipes in these motors that are not a single cast piece: the larger cast pieces have extruded inlet/outlet tubes that are connected using an adhesive. There is no metal-to-metal friction or press-fit to keep these tubes in place, so after enough heat cycles the adhesive will soften/loosen up and the tube will come out of the cast block (with the hose still attached), resulting in a rapid loss of engine coolant.

Regardless of how the vehicle is driven, it seems this problem may eventually effect all 996/997 Turbo, GT2 and GT3 models including the 2010+ GT3 and GT3RS.

We have heard of many cases of this, and in fact the last time we were at Infineon Raceway, Alex and I were talking to someone about it when the exact problem happened to a 997 GT3 right in front of us -- a large steam cloud evacuated the rear of the GT3 and it spun on the hairpin! Luckily he missed the other cars and the barriers.

Here's a video showing how the tube comes apart:

Here's a diagram showing where the problem happens on 997 GT3 models:


Here's a picture from a 996 Turbo's coolant pipe. Both of these tubes you see are slip-fit into the cast piece and fixed in place using an adhesive:



We first saw this in early 2006 when we started building 996 Turbo engines. In fact, our 996 GT3 suffered from this exact problem earlier in its life.

James has a technique for fixing this, something we do for all 3.9L motor builds and any time a GT1 motor is removed from the car.

This week we had a 2011 GT3RS (project that was documented earlier) whose lucky owner will have plenty of track time with her. He decided that rather than wait until it might happen, he would rather be on the safe side and fix the problem now, preventing a scenario where the coolant might cause a high speed crash at a local track.

The bumper is removed exposing the Shark Werks GT3 Bypass Exhaust:


And despite the car's super low mileage, its factory fresh and perfectly running motor is removed from the car:





With the motor removed from the car, James locates each of the potential problem areas and drills a small hole through the cast piece and into the tube (that normally would come flying out at some point in the future). Then he threads a bolt through both pieces and using loctite secures the bolt in place. This locks the tube together, preventing it from coming apart regardless of heat cycles.





The coolant pipes will now act as a single piece.


All the tubes are buttoned up...



And the motor is lifted back into place, the remaining details addressed.



And here she is, now ready for an enjoyable weekend around the track at 8000 RPMs.




For more of this car, check out the whole project here.

raineycd

Great info! Can the fix be done with the engine in? Looks like it can be done removing the alternator?

How did you keep metal out of the pipes when drilling?

Thanks!!!!

jpflip

I don't know about your fix. My concern is due to the fact Porsche used a wrong type of glue, this glue will eventually dry and break and even if your screw is holding the pipe it will leak between the pipe and the casting ???

SoloRacer

My thoughts exactly. The engine is out of my car to install a clutch/LSD and to do fix the coolant lines. I'm having a set of CNC machined sleeves tig welded in to permanently fix the problem. Considering the expense of dropping the engine it's not that much more to weld the sleeves in vs the bolt/screw method. Is there an advantage to the bolt/screw method that I am missing?

Mikelly

I echo the above two posts and thought the exact same thing last night when I read this...My car actually had the suspected parts replaced when I bought it back in 2007, while the dealer had it. But I'm still concerned... Might have to go the "tig welded" route...

Mike

LVDell

I'd trust a welded solution much more than leaving the same old hoses with the same old adhesive. Besides, is the screw and loctite solution tested to the rigors of coolant temp and pressure? Is loctite permeable?

blockhed

i'm also interested in this as i fear for this to happen as my car has 33k miles and it's been driven since I have owned it.

Kevinmacd

Excellence mag, also pointed out the joint problem, and showed that tig welding around the joint was a better option.

jpflip

I got no affiliation with the seller but you can get the fittings from him !!!


http://cgi.ebay.com/ebaymotors/ws/eB...T#ht_668wt_941

PAULUNM

He says there are 6 fittings- I thought there were 8? Can anyone confirm?

jimq

I would assume as usual Porsche says this is not a real problem and wont pay to fix it?

LVDell

Under warranty of course.

Dan@SharkWerks

Some of the fixes can be applied with the motor in the car, but for all of them it would need to be removed or disassembled. We find it's easier to remove the motor in this case to do it right.

Since most of the drilling (until the hole punches through the other side) happens on the outside of the motor the chips are going out and we clean them with a shop vac. As the bit goes through to the inside, we vacuum from the inside. There is minimal if any significant metal chips left by this point, not enough to effect the cooling system or damage anything.

We have not seen coolant leak through the adhesive yet, it seems that's not the issue necessarily, there's plenty of adhesive and it does not dissolve (it softens and acts like an o-ring to seal between the tube and casting), it just won't hold that tube in place anymore.

Mike,
Those replacement parts will not be a permanent fix. If the dealer simply used more OEM parts with this same adhesive - coolant pipe junction, it will probably have the same issue down the road.

The main benefit to the screw method is it's less complicated. Welding the components may be as effective under some circumstances (good welding, machined replacement fittings) but if it's poorly welded it won't or will leak. It's more work/money and the motor must be further disassembled to allow access to weld in there etc. Either solution done properly should fix the problem.

We have used this method on at least 16 motors we have built here since 2006, none have had the problem since. From Loctite's page on the Red variety: ] There is a full manual on specifications there but I have not checked it out (http://www.loctiteproducts.com/p/t_l...er-Red-271.htm), I believe it goes into how Loctite red may breaks down etc, this is something James here would know better than I would.

Excellence does document both our method (see Option 4.) and the TIG method (see Option 1.) in the February 2011 issue. They do not recommend the machined fittings for Option 1 (mentioned above and in other threads) -- I would recommend machined fittings if welding, they will work much better than the existing tubes.

The only worry in the Excellence article with Option 4 is that the coolant will leak through the threaded/loctite bolt. We have not seen this problem and I do not believe the Loctite will fail under these circumstances.

Eureka! Mentioned above, this seems like the way to go if you're welding

The GT1 motors have different numbers depending on the application/vehicle, I can't confirmed what's included in that Ebay auction without testing it though.

Mikelly

Dan, my comment wasn't meant as if the issue has been permanently addressed. It's meant that it was an issue that was dealt with once, but in the "OEM" manor. I'll probably buy the collars and have the housing repaired properly with welds...

Mike

Dan@SharkWerks

Hi Mike, ah okay - yeah I was simply sharing a warning. That's interesting they would fix it with the same parts that cause the problem in the first place.