Hi
I am not sure I understand the issue you wish to address. If you are bolting a cover to some pressure vessel, then the following should be considered :
Internal pressure acts on the cover and produces tension force in the bolts.
A certain clamping force is required on the gasket to prevent it sliding out.
Bolt tension must be at least equal to the sum of these forces.
Bolt/nut thread friction and nut/washer friction must be allowed for in calculating tightening torque, or use some load indicating method.
Hi Richard...
Do I use the equation Pressure x Area to determine the load on the bolts...then divide this number by the number of bolts?...if so, then what information do I need to know to determine whether the bolts themselves can handle this load?...
I'm bolting two flanges together with an O-ring type seal...approximately 1000 psi internal...#10-32 bolts, 140 ksi ultimate (3 bolts w/nuts)...
Pressure x area gives force on all the bolts. Dividing by number of bolts gives force on each bolt. Divide force on each bolt by 0.0200 square inches. (This is the tensile stress area for a #10-32 UNF bolt. One source for this information is the Machinery's Handbook; find one and verify the number I quoted you.) The result is the pressure on each bolt, and needs to be lower than the yield strength of the bolt as a minimum. (Don't use ultimate as you show above) How much lower depends on the factor of safety that you are designing with for a high pressure system.
When the bolts are torqued, they put the O-ring under maximum compression (total tension in bolts produced by torque divided by length and width of O-ring).
When the vessel or pipe is pressurized, the compressive force on the O-ring is reduced by the force that tries to separate the two parts being connected.
A good design will determine the total preload force in the bolts necessary to keep the minimum compression on the O-ring when the vessel is pressurized to be at least the minimum value recommended by the O-ring manufacturer or the Pressure Vessel Code. This is required to avoid leakage.
Notice that the pressure inside the vessel does not change (theoretically) the tension in the bolts. It just reduces the compression on the O-ring.
Normally, the bolts are preloaded to around 90% of their proofload. Verify this on the Pressure Vessel Code.
AJohnson...
thanks!...i understand what you are saying...i think that was the missing link i was looking for...(the tensile stress area)...this eng-tips is great stuff!...
I would to Section VIII-1 of the ASME Boiler and PRessure Vessel Code and check the design against the rules in Appendix 2. A pressure of 1000 psig is no small matter and the design needs to be right. The use of 3 very small bolts (machine screws really) in this kind of pressure service does not feel right to me at all. I deal with these kinds of pressures on a daily basis and would not dream of doing this type of design on the fly without checking against the ASME Code. You need to be very careful with high pressure service.
Chris,
It seems like all this analysis ASSUMES the bolt torque method of tightening, gives repeatable and accurate bolt tension. Yet no Mechanical engineering professor would recommend using a torque method to achieve 90% of yield preload tension, because 80% to 90% of the torque is nothing but friction. And no matter what torque method you use or lubrication, this friction varies 25% or greater. Using a perfect torque wrench on the same bolt still gives 25% or greater variations in bolt tension. Then when you are tightening a flange bolt the stiffer the flange joint, the greater the bolt tension variations (50% to 200% errors). If you could come up with an inexpensive way to measure bolt tension more accurately, you could become rich. For flanges, see the actual typical graph of bolt tension variations:
