Bolts in Shear on Circumferential Bolt Pattern

[sleepdrifter]

I'm designing a part that will attach to a circumferential bolt pattern, with threaded holes on the mating component. In my attached picture below, the blue components is my part that will have thru holes, the orange component is the mating component with threaded holes, and the grey component will be shims I use to occupy the .020" gap I'm leaving to put the joint into friction. The mated assembly will be sealed and pressurized to 15 psi.

My primary question is, can I just use a simple bolted joint shear calculation on this? I was referencing Bolted Joint Analysis 4th addition, section 19 for joints loaded in shear. And using a grade 8 3/8-24 fastener and the pressurized diameter being 10.7 inches, I got a decimal when solving for the number of bolts I needed to withstand a force of 1348.8 lbf. I know it's not a massive force but I feel that my calculations are off. Can someone help point me in the right direction of how I can solve for number of bolts needed considering my joint? I believe I oversimplified the calculations.

 

[SWComposites]

you need to check:
-fastener shear
-bearing in orange part
-bearing in blue part

what are you using for fastener shear strength? part bearing strengths?

 

[Stress_Eng]

At a first glance, the bolts will be unsupported over the shim thickness. Internal pressure will put the bolts into shear, and, due to the unsupported length, I'd suggest a bending calc'n. Due to internal pressure, the horizontal sections of the orange and blue parts will tend to bow radially outwards and inwards respectively. In addition, the vertical walls will also see pressure loading, and again want to bend (also rotations at intersections with horizontal members). All this bowing / bending may put the bolts into tension, due to vertical wall bending and prying [bolt tension to hold the parts together and abutment contact (prying)]. The bolts will see shear, bending and tension. All the loading is from the pressure and the pressure load is dictated by the surface area. Your calculations will therefore need to be based on an angle (r.theta), and two opposite bolts will be in the middle of this length.

 

[MintJulep]

It's not a very good design.

Hard to assemble, and will almost certainly leak.

 

[mfgenggear]

I'm designing a part that will attach to a circumferential bolt pattern, with threaded holes on the mating component. In my attached picture below, the blue components is my part that will have thru holes, the orange component is the mating component with threaded holes, and the grey component will be shims I use to occupy the .020" gap I'm leaving to put the joint into friction. The mated assembly will be sealed and pressurized to 15 psi.

My primary question is, can I just use a simple bolted joint shear calculation on this? I was referencing Bolted Joint Analysis 4th addition, section 19 for joints loaded in shear. And using a grade 8 3/8-24 fastener and the pressurized diameter being 10.7 inches, I got a decimal when solving for the number of bolts I needed to withstand a force of 1348.8 lbf. I know it's not a massive force but I feel that my calculations are off. Can someone help point me in the right direction of how I can solve for number of bolts needed considering my joint? I believe I oversimplified the calculations.

I would recommend re thinking this design.
Why not use flanges?
Or what is the intent of this assembly.
Might be better with a weld joint. Or make
The joint a slip fit. With a low temperature braze.

 

[goutam_freelance]

You have the option for a lower size and/or lower grade of bolts.

However, a good factor of safety will take care of bending stresses.

 

[Orange_kun]

As others have said, I'll tell you right now... that won't hold pressure for long.

It wont't be super round, I haven't heard any mention of an elastomeric sealing element, and I don't see how you're sealing the bolt clearance holes.

I would recommend looking up a design guide for face sealing flanges. Way more robust sealing and easier calculations for bolt forces.

 

[sleepdrifter]

you need to check:
-fastener shear
-bearing in orange part
-bearing in blue part

what are you using for fastener shear strength? part bearing strengths?

For fastener shear strength I'm using Grade 8 bolts, so 60% of it's tensile strength which is 90ksi.
Not sure how to go about the part bearing strengths. From what I understand, I would need to consider failure of the bolt and failure of the plates.

For failure of the bolt I can use a simple shear stress equation, using my total force/# of bolts and the cross-sectional shear area of the bolt.
I'm struggling to find an equation to calculate failure of the hole material itself.

 

[sleepdrifter]

At a first glance, the bolts will be unsupported over the shim thickness. Internal pressure will put the bolts into shear, and, due to the unsupported length, I'd suggest a bending calc'n. Due to internal pressure, the horizontal sections of the orange and blue parts will tend to bow radially outwards and inwards respectively. In addition, the vertical walls will also see pressure loading, and again want to bend (also rotations at intersections with horizontal members). All this bowing / bending may put the bolts into tension, due to vertical wall bending and prying [bolt tension to hold the parts together and abutment contact (prying)]. The bolts will see shear, bending and tension. All the loading is from the pressure and the pressure load is dictated by the surface area. Your calculations will therefore need to be based on an angle (r.theta), and two opposite bolts will be in the middle of this length.

Yea I'm seeing that now.. I'm not using a typical friction joint here and realize the bolts are just being put in bending. And I agree that the shims really aren't going to be doing much for me except reducing any play that may be present.

Thanks for the advice

EDIT:

I've attached a rough FBD of the forces, can you let me know if my simplification is accurate? I'm seeing shear primarily from the blue plate as the fastener is fixed into the thread on the orange (but a shear force is present on the orange as well). Torque stress due to the torque applied to fasten the bolt (I'd like to simplify this portion if possible). And a bending stress due to the fact that my plates are not forming a friction joint and there is a gap.

 

[sleepdrifter]

It's not a very good design.

Hard to assemble, and will almost certainly leak.

I would recommend re thinking this design.
Why not use flanges?
Or what is the intent of this assembly.
Might be better with a weld joint. Or make
The joint a slip fit. With a low temperature braze.

As others have said, I'll tell you right now... that won't hold pressure for long.

It wont't be super round, I haven't heard any mention of an elastomeric sealing element, and I don't see how you're sealing the bolt clearance holes.

I would recommend looking up a design guide for face sealing flanges. Way more robust sealing and easier calculations for bolt forces.

I didn't include all the details of the design because it didn't really seem pertinent to my question. There's going to be an o-ring sealing the two parts.

I'm forced in using the orange part with a circumferential bolt pattern and it is not my design. I'm only here asking this question because it's my first time having to design a pressurized fixture for a circumferential bolt pattern and haven't had to deal with calculating the forces in this scenario before.

 

[Orange_kun]

There's no way to make a flange that gets bolted vertically to back up the orange cylinder? Why are you being forced to use it? What is this device doing?

I have never encountered a bolted joint like this. Put threads in a cylindrical surface, whereby the fasteners all compete to distort the female threads? Doesn't seem ideal to me.

If the system is pressurized and sealed, I'd be more inclined to run a hardened shaft thru all the components in double shear. Have the sealing o-ring below it, and another above to hold things roughly concentric. Even a 1/4 IN pin/rod should be fine for your nominal 1400 lbf.

 

[sleepdrifter]

There's no way to make a flange that gets bolted vertically to back up the orange cylinder? Why are you being forced to use it? What is this device doing?

I have never encountered a bolted joint like this. Put threads in a cylindrical surface, whereby the fasteners all compete to distort the female threads? Doesn't seem ideal to me.

If the system is pressurized and sealed, I'd be more inclined to run a hardened shaft thru all the components in double shear. Have the sealing o-ring below it, and another above to hold things roughly concentric. Even a 1/4 IN pin/rod should be fine for your nominal 1400 lbf.

I pressure test customer components in the aerospace industry. This test is just to validate a seal (not related to this interface I'm showing). The actual usage of this circumferential bolt pattern is unknown to me, and the pressurization of this part doesn't depict the real use case of this weird bolt pattern.

I agree using a pin of sorts would likely be better and is likely how I assume the customer uses this bolt pattern. But I'd like to do that math on if just using off the shelf bolts for 5 minutes at this low of a pressure will suffice. I don't care if I need to replace the bolts every test, it's pennies to creating an expensive and overdesigned fixture if I can help it.

I have 56 total holes to use over a 10 1/2" diameter bolt pattern (there's two sets of circumferential bolt patterns with 28 holes each). I feel if I utilize at least the first set of 28 I'd be fine with the force I'm seeing, but I want to do the math to validate.

 

[Orange_kun]

Honestly, this just seems like such an odd way of retaining it. When designing test fixtures, I've never been told I HAVE to use part features, unless those were the features specifically being tested. You're trying to design a rather complicated joint for... why? To use features of a part that aren't related to the test?

I'd go with close fitting pins with a shoulder to prevent them falling inside the cylinder. Close sliding fit in the fixture holes, clearance in the orange part. Easy and quick to slam 4 or 8 into a row of holes, use a rubber band to retain them. Test. Choose a relatively soft material if the orange part is delicate. Or if you're worried about distorting the orange part's threads, fine, get some shoulder bolts. But I don't think I'd be trying to actually develop clamp load in the bolts. Then you'll be faffing around with shims and tightening bolts and hoping you don't distort the orange part's cylindricity. And I still have misgivings about using the orange part's threads at all for putting it under test.

But if you're dead set on it, the math already showed you a single bolt is enough to withstand the shear (was that calc in single or double shear?) so the risk is really fatigue, because you're bending threads, or damage to the orange part by using less than the 56 holes to withstand the force. I don't know how concerned you need to be about that. If it's aerospace, it might not be ideal to load up a low number of those holes with the pressure load. Is it steel? Nylon? Nitronic 60? Unless I'm misunderstanding which part is actually under test and the orange part is just another existing fixture component.

The difficulty here is not how many bolts you need, but rather what you're allowed to do to the parts, how long the test setup takes, and how many units need to get tested. If none of that matters to you, then just use all 56 holes finger tight and replace the fasteners every time.

 

[Stick.]

Just focusing on your original question, I don't think you're missing much. If you just look at shear in the bolts, assuming 28 bolts, the shear stress is about 0.7 ksi per bolt. Yes, there will be some bending stress, but since your loads are low and the span is small, you have to ask yourself if the extra time doing the calcs is worth it. I probably wouldn't worry, but I could be missing something when the problem is abstracted over the internet. My take on your questions about bearing stress in the parts is pretty similar, unless these things are made of something very soft or are fragile/delicate.

 

[sleepdrifter]

Honestly, this just seems like such an odd way of retaining it. When designing test fixtures, I've never been told I HAVE to use part features, unless those were the features specifically being tested. You're trying to design a rather complicated joint for... why? To use features of a part that aren't related to the test?

I'd go with close fitting pins with a shoulder to prevent them falling inside the cylinder. Close sliding fit in the fixture holes, clearance in the orange part. Easy and quick to slam 4 or 8 into a row of holes, use a rubber band to retain them. Test. Choose a relatively soft material if the orange part is delicate. Or if you're worried about distorting the orange part's threads, fine, get some shoulder bolts. But I don't think I'd be trying to actually develop clamp load in the bolts. Then you'll be faffing around with shims and tightening bolts and hoping you don't distort the orange part's cylindricity. And I still have misgivings about using the orange part's threads at all for putting it under test.

But if you're dead set on it, the math already showed you a single bolt is enough to withstand the shear (was that calc in single or double shear?) so the risk is really fatigue, because you're bending threads, or damage to the orange part by using less than the 56 holes to withstand the force. I don't know how concerned you need to be about that. If it's aerospace, it might not be ideal to load up a low number of those holes with the pressure load. Is it steel? Nylon? Nitronic 60? Unless I'm misunderstanding which part is actually under test and the orange part is just another existing fixture component.

The difficulty here is not how many bolts you need, but rather what you're allowed to do to the parts, how long the test setup takes, and how many units need to get tested. If none of that matters to you, then just use all 56 holes finger tight and replace the fasteners every time.

I agree with everything you stated, and I don't have to use the provided part features, but this is the only feature on this entire part for fastening. I could develop a fixture that clamps onto the customers part, but then I risk damaging/marring it up.

The customer part is a titanium alloy. I'm likely going to design my fixture using 4130, potentially with a heat treat to keep the clearance holes from distorting.

 

[sleepdrifter]

Just focusing on your original question, I don't think you're missing much. If you just look at shear in the bolts, assuming 28 bolts, the shear stress is about 0.7 ksi per bolt. Yes, there will be some bending stress, but since your loads are low and the span is small, you have to ask yourself if the extra time doing the calcs is worth it. I probably wouldn't worry, but I could be missing something when the problem is abstracted over the internet. My take on your questions about bearing stress in the parts is pretty similar, unless these things are made of something very soft or are fragile/delicate.

Thank you for helping me validate that original question. I am just looking for enough validation in simple calcs to ensure I don't need to dig further into this. The bolts can be consumed after the test for all I care.

Orange component is a titanium alloy so it's plenty strong and my fixture will likely be a hardened 4130 or something similar.

 

[mfgenggear]

Just focusing on your original question, I don't think you're missing much. If you just look at shear in the bolts, assuming 28 bolts, the shear stress is about 0.7 ksi per bolt. Yes, there will be some bending stress, but since your loads are low and the span is small, you have to ask yourself if the extra time doing the calcs is worth it. I probably wouldn't worry, but I could be missing something when the problem is abstracted over the internet. My take on your questions about bearing stress in the parts is pretty similar, unless these things are made of something very soft or are fragile/delicate.

Early in my career, I tested large pipe. 1000 PSI.
Hydrostatic testing. A large fixture with pots at each end . Inside where o ring grooves . With a gantry the pipe was lifted. One side slide into the left hand side. The other end was hydraulic slid into place. Then tested for leaks. In your case
You need to make a similar fixture to retain
The assembly in compression.
The radial preload of the bolts may distort the assembly. Need to test as it will be assembled.
That the test will fail. It can't be any clear than that. You need to discuss with a senior member or supervisor.

 

[SWComposites]

I'm struggling to find an equation to calculate failure of the hole material itself.

its a simple bearing stress calc for each plate: Fbr (psi) = single fastener load (lbs) / (diameter * thickness)
allowable bearing stresses can be found in MMPDS.

 

[mfgenggear]

So this bolt assembly considered screw Assembly?
Preload and stiffness washer diameter of the washer or washer head of the screw configuration will effect the stiffness .
Which requires clamping of the members
Friction from the preload will then retain the clamping force. Correct?

 

[mfgenggear]

PS I wanted to add , the OD of the pipe will require a spot face. To have proper contact with the Bolt washer , so that it is making full contact.
The washer will be contacting on high spots.
Making preload incorrect.