Propane Vaporizer/Regulator leaking Gasket blowout

[MCasey]

O.k. I'm sure this will be an easy question for you guys based on reading all the posts....

I have a small propane vaporizor that is experiencing gasket blowout.

The device is about 5" in diameter. It has a coolant passage where engine coolant passed though to provide heat. It has vapor passage where the vaporized propane passes though. We are having aproblem with the gasket blowing out between the coolant passage and the propane passage. Coolant passage sees 40PSI of 160 F coolant. Propane passage sees 5 psi 90F. The gasket is nitril reinforced with nylon. We use (6) 1/4" screws to secure the cover which clamps the gasket. We torque the screws with 48 in-lbs of torque at assembly.
On warranty returns we measure the break away torque on these screws and we see torque as low as 1 in-lb on a screw up to 20 in'lbs (very random what we are seeing).
It certainly appears to me the 48 in-lb of torque is way too low. How can I calculate the correct torque for this design? (When I asked the engineer responsible for the design he said "it's what we use on the other design" (even though the designs are completely different) Scary.
Any help or direction would be greatly appreciated. I've also uploaded a picture of the gasket blowout as well as what the device looks like.

Thanks!

 

[chicopee]

Were the six bolts torqued evenly during reassembly with a torque wrench?
As far as the torque of 48 in-lbs may be reasonable depending on the metal composition of the vaporizer and bolts.
Some typical values of SAE 1/4" bolts grades 1 thru 8(lightly oiled) torque values range from 5 ft-lbs thru 10.5 ft-lbs whereby the internally threaded body would be steel or cast iron.
If the composition of the vaporizer is aluminum than obviously torque would be less than those values presented above.
Get your hands on the Machinist Handbook and you'll find numerous formulae for torque calculations.

 

[chicopee]

One more thing, is a torquing sequence advised by the manufacturer?

 

[MCasey]

The 6 bolts are torqued 'correctly' during assembly. All bolts are torqued in a star pattern at 10 in-lbs and then at the 48 in-lbs in a star pattern multiple times.
If it is dis-assembled 24 hours later residual torque is down to 25 -in lbs. (one of our issues is that when we first noticed this issue as we noticed our process uses dry screws no lube. We reccommend using lube, but the decision was made just to use a higher torque. (we used to only use 36 in-lbs of torque)
The body is cast aluminum. I have been able to apply 110 in-lbs of torque before the screw heads cam out. I would like to increase the torque to 75 in-lbs but I don't know what the 'crush rating' of the gasket is. (manfacturer of the gasket has no clue what this is) The gasket is nylon reinforced nitril sheet gasket.

We are the manufacturer of the device so we are the ones specifying the torque. (Again...scary....)

 

[ApC2Kp]

MCasey,

Propane liquid 90 deg F would have vapor pressure 150 psig.

Liquid propane at 5 psig would be -30 deg F.

Your device could be seeing much higher pressure, or hi-lo temperatures that could result in torque relaxing.

The engine coolant pressure and temperature seem typical for automotive services.

The residual torque 25 in-lbs from original 48 in-lbs would seem to indicate that perhaps thread lube is needed for more accurate initial torque on bolts. If assemblers don't like the mess of thread lubes, maybe a dry lube could be used in the aluminum threads? Maybe a thread locking compound is needed for a propane application (tamper-proof). Or the gasket material could be relaxing that much. Try a RTV silicone seal compound instead of gasket?

Most house installations have dual regulator setup for flow from tank regulator 3-5 psig thru reasonable pipe size to the 2nd regulator at house connection and 10-15" w.c. house appliance service.

Is this vaporizer / regulator used as single pressure drop device in your system? If liquid propane is fed to this unit for vaporization, then there could be vapor plus liquid in the inlet to your vaporizer. That is a difficult job for a regulator to function on mixed liquid and vapor flow.

 

[chicopee]

If the bolts are of steel,you would have two different coefficient of expansion. Aluminum expands a lot more than steel, therefore, stressing the bolts perhaps past their yield point reducing the initial torque to their residual value or the threads in the aluminum block have slightly deformed also reducing the pre-stress torque.

 

[chicopee]

Also,recheck the length of engagement of the treaded bolts. With two dissimilar material (where bolt is steel and part is aluminum)that length of engagment must be increased.
light Lubrication of the threads mentioned above is also a good point to prevent galling of the aluminum threads.
An excellent reference is the Machinist Handbook