jetmech From Australia, joined Mar 2006, 2586 posts, RR: 53 Posted (3 years 3 months ago) and read 4648 times:
G'day Techies,
JetMech has been pondering again, this time, about the fatigue life of bolted joints and the effect of bolt pretension. Pretension of bolted joints is important especially where the bolt takes tensile loads, and the joint members are compressed, such as engine mounts.
The idea is to place a preload on the bolt which exceeds any foreseen load attempting to separate the joint members. Doing so reduces the magnitude of the cyclic loading on the bolt, which in turn, greatly extends the fatigue life of the bolt. Where I am getting confused however, is whether bolt preload - prior to joint separation - eliminates the bolt from "feeling" any of the external load, or whether it reduces the magnitude of the external load felt.
The "spring" model used to explain the behaviour of bolted joints seems to indicate - to me at least - that preload prevents the bolt from feeling any of the external load prior to joint separation. In other words, any external load - with a magnitude less than the preload - applied exactly equals the reduction in joint clamping force, the net effect on the bolt being zero.
However, if we take a more detailed look at joint behaviour using joint diagrams, we see that the bolt indeed "feels" part of the external load, albeit with a greatly reduce magnitude which itself depends on the relative stiffness of the bolt and joint members.
rwessel From United States of America, joined Jan 2007, 1988 posts, RR: 2 Reply 1, posted (3 years 2 months 4 weeks 1 day 20 hours ago) and read 4587 times:
Using the number from the first diagram: so long as the external load is less than 1000lbf, the bolt will feel a load of 1000lbf. But the *source* of that load varies. If you have a 250lbf load, the bolt will feel 1000lbf, 250lbf from the load, and 750lbf from the block. With a 900lbf load, the bolt will feel 1000lbf, 900lbf from the load, and 100lbf from the block. With no load at all, the bolt will still feel 1000lbf, but all of it from the block.
jetmech From Australia, joined Mar 2006, 2586 posts, RR: 53 Reply 2, posted (3 years 2 months 4 weeks 1 day 3 hours ago) and read 4494 times:
G'day Rwessel,
Sounds good. I was originally not intuitively sure how pretensioning benefited fatigue life at all, as the way I originally saw it, any additional external load had to be "felt" by the bolt in full as it was the only member holding the joint together. However, I slowly realised that the key to it all was the clamp load developed in the joint members, which were placed under compression by the bolt.
Thus, any external loading applied to the joint tending to separate the members, acts in a manner to relieve some of the clamping force between the joint members. In theory, this reduction in clamp force will be equal to the external load, so the bolt will not "feel" any change in tension prior to joint separation.
This is in line with what you mentioned, and is what intuitively leads me to believe that the bolt feels none of the external loading prior to joint separation. However, the joint diagrams which also make sense to me say that the bolt does indeed feel the external load, albeit at a greatly reduce magnitude, Herein lies my conundrum!
Regards, JetMech
JetMech split the back of his pants. He can feel the wind in his hair.
tdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 81 Reply 3, posted (3 years 2 months 3 weeks 5 days 18 hours ago) and read 4320 times:
Quoting jetmech (Thread starter): So what actually is occurring? Prior to separation of the joint members, does the bolt feel none or some of the external load?
It feels some.
The top concept is based on a very stiff block...i.e. modulus of the clamped material is much much higher than modulus of the bolt. In the limiting case of an infinitely stiff block, the bolt feels nothing until you exceed preload. This is equivalent to the joint diagram where the slope of the green line (stiffness of the clamped material) is vertical. If the clamped material has some stiffness, there's sharing of loads in proportion to the ratio of stiffness.
However, in this particular example, I think the loading is different in your two cases. In the top case, you're loading the bolt itself. In the bottom case, you appear to be loading the clamped pieces.
jetmech From Australia, joined Mar 2006, 2586 posts, RR: 53 Reply 4, posted (3 years 2 months 3 weeks 5 days 2 hours ago) and read 4244 times:
G’day Tom,
Thanks for that! Many of the other sources of material I have looked at also say that the bolt does feel some of the external load, though at a much lower magnitude.
Quoting tdscanuck (Reply 3): modulus of the clamped material is much much higher than modulus of the bolt. In the limiting case of an infinitely stiff block, the bolt feels nothing until you exceed preload. This is equivalent to the joint diagram where the slope of the green line (stiffness of the clamped material) is vertical. If the clamped material has some stiffness, there's sharing of loads in proportion to the ratio of stiffness.
I should have realised that the situation of the bolt feeling zero load was a limiting case of this general theory. As I recall, the hardest aspect of designing bolted joints is establishing the exact stiffness of the members being clamped.
Regards, JetMech
JetMech split the back of his pants. He can feel the wind in his hair.
Sponsor Message: 
From Australia, joined Mar 2006, 2586 posts, 
From United States of America, joined Jan 2007, 1988 posts, 
From Canada, joined Jan 2006, 12709 posts, 
Sponsor Message: 
Printer friendly format
