Bolted joint with gasket

[APH]

I have a molded plastic part nut, piece of aluminum stamped part, a gasket, and thread forming screws. This seems to be pretty straigth forward stuff. I need to figure out what clamping force is needed to seal up the joint. This clamping force needed to withstand various vibration effect, etc. I also need to figure out what torque requirement to produce this force and to see if this would also strip out the plastic thread (produced by thread forming screws).

Anyone has a web site or step by step formula, so I can check my calculation.

Thanks
APH

 

[Screwman]

You have an extremely complex system to try to analize. Both the gasket and the plastic component will have somewhat nonlinear load/ deflection curves which is going to make it even more complex. When we do an analysis like this which has two "plastic" components, I figure that it will take about 40 man hours, including doing the load / deflection testing of the components (don't trust what you get from the gasket companies).
A more time effective method might be to tighten the screws up to failure and then take 75% of that as a seating torque and try one of the assemblies in test. You will have to determine the cost benefit ratio for this joint to determine how much time to invest in the joint analysis.

Dick

 

[APH]

I see your point there. I was just trying to start from somewhere.

APH

 

[arto]

You may want to take a peek @ some of the analysis in ASME B&PV Code Sec. VIII DIv.1 App.2 for flange design.

Basically you need to develop enough contact force to seat the gasket - used to be 1600psi for a generic "y" value - check the gasket mfrs catalog [Garlock has good info

http://www.garlock.net/

]- O-rings have 0-200psi to seat; & also take the pressure force once it's sealed.


If it's too much force for the plastic part, maybe use Helicoils or Keenserts.

You could get the req'd torque from the good old F=5T/d formula in Shigley.

 

[APH]

Thanks guys,

My application is not like vessel design and stuff. This is an electronic component ans has no internal pressure except from vibration and atmoshperic condition. I could still use the "Y" value, but I'm having problem with finding out the parameter that I need to use "Y" value. One of the key value is to determine minimum seating stress for the gasket. this would require me to know the stiffness of the material. I'm having a hard time to find this, unless getting it from experiment which I'm going to do anyway.

APH

 

[CoryPad]

Can you design the gasket to be an embedded type so that when it is compressed fully, you have solid contact between the clamped components? This type of joint usually has a much higher joint preload than the gasket force, so you can ignore the gasket force when determining the tightening torque. Since you have a prevailing torque feature (thread forming screw), the torque will need to be higher than that estimated by T = 0.2 F d.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[APH]

Cory,

Basically you're saying an O-ring type gland. Unfortunately, it's pretty late in the game to do this. THe "0.2" constant on your formula. Is that the K factor for zinc finish type on the screw?

APH

 

[Screwman]

The .2 that Corey is using is the K factor, but it really will be higher for a thread forming screw into plastic. The two big variables in this joint are the gasket and the nut member. Trying to maintain clamp load in threads that are tapped into plastic gets to be a real problem because of the amount of creep that many resins experience when subjected to the pressures put onto them by the threads. This is very dependent on the resin and filler package that is being used. We have seen and average of over 20% clamp load loss in plastic joints after 24 hours and when a thremal cycle up to 100 Degrees C is added, the clamp load loss has increased to over 80%.
The use of a self energizing gasket like an O ring will greate a seal while requiring almost no clamp load.

Dick

 

[APH]

How much higher does the K factor of the thread forming screw get?. The nut material is Valox. Gasket is Nitrile durometer 60, and cover is Aluminum, and the tread forming screw is Plastite 4-40 x 3/8.

APH

 

[Cockroach]

Yeah, I agree with all the above. I would of not used a gasket in this application since sealing performance depends heavily on "clamping pressure" which is induced from your screws on the flange bolting.

O-Rings certainly get around this, even in face seal applications. Since compression carefully controlled, seal is achieved well below the torque requirements that could prematurely strip out the box threading of the flange.

At this stage of the game, redesign and manufacturing is out of the question, I would be looking toward a gasket adhesive which would promote sealing properties at relatively low clamping pressures. There are several on the market which work well on plastic surfaces. The advantage here is that your flanging is plastic, but be sure to look at the adhesive chemical properties since many can compromise the plastic surface.

Good luck with it!

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[APH]

I've built a prototype gasket joint on this issue and I sprayed it with 2000 psi 4 gallon a minute water. I've used Nitrile rubber with adhesive on the back. I got mixed result. It seems like they are not consistent. I have 5 assemblies made and tested with the same parameter, 3 of them failed, 2 of them passed. Hmm...I;m stuck

APH

 

[Tmoose]

I'd put hard spacers/standoffs between the flanges. One around each bolt. Then apply silicon to create a formed-in-place gasket and torque it up. Or, make them a bit shorter than than your Nitrile gasket. Asking a clamped gasket to be part of even a minor structural joint has brought me trouble many times.

On the stuff I've worked on In general flange thickness/stiffness has been under-estimated.
If the flange is "thin" and the bolts "widely separated" or too far from the main part diameter or structure the the clamping pressure gets real low between the bolts. The joint looks like this, ^|^|^| .

 

[APH]

I put pressure sensitive film and you are right. The pressure in between bolts are a lot less. Since I can't change the total thickness between the cover plate and gasket, I'm planning to use less durometer gasket with stiffer cover plate hoping that the pressure is translated better between bolts. In doing the gasket shape itself, does it matter to have a thru hole or a slot for bolt diameter clearance?

APH

 

[Tmoose]

Do you have room behind the weak flange to add some reinforcement plates? For instance , 2 "C" shaped plates with the bolt pattern added.

IN extreme cases 1 plate per bolt can be used.

http://www.cantonracingproducts.com/valve_covers/65000.jpg

 

[CoryPad]

APH,

In almost all cases, a hole is preferred to a slot.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[APH]

I have to look at the assembly first. It looks like I'm pretty limited on the overall flange thickness. The "C" cross section will increase the "I" value. I have about another .100, which is the heigh of the screw head, otherwise it will stick out too far and interfere with other parts.

APH

 

[APH]

CoryPad,

Could you shed more light on your respond.

Thanks
APH