Calculate Max Axial Load for Large Threaded Inserts

[jaap de vries]

Dear Y'all

I have a large pressure vessel 12'' ID 1.5'' wall that I use for combustion experiments (UCF). The cylindrical vessel has two inserts on each side that are attaches through large threads (12''-4 UNC), 12 inch nominal diameter an 4 threads per inch, about 2.5'' deep. I want to calculate the maximum axial load (pressure in vessel) the material is 7075 Aerograde aluminum. Hope someone can help me.

Kindest Regards, Jaap

 

[Q_Goest]

Hi Jaap,
Can you provide a drawing for this? Please show as much detail as possible. Also, what temper is the 7075? Is there any cyclic stress on it?

 

[FredGarvin]

12-4! Wholly molley. That's a big thread. Cool!

To add more questions, since this is a combustor test, what is the expected temperature range? Even the 7000 series has a pretty steep temperature degradation curve.

Also, are these threads manufactured by a company like Rosean or Fairchild or are they your manufacture? How are they inserted and locked in place?

 

[jaap de vries]

More Info

The threads are standard Unifified Course. It is not a combustor. I will monitor a freely propagating flame using a High speed camera so I don't expect high temperature effects (only for a very short time). The design was not mine but I had it made.
Threads are made by a standard (good) machinist. Since the ASME boiler and pressure vessel code manual does not advise such large hreads it is hard to find inormation. I just need to know within a certain factor what strength to expect.
See attached PDF

 

[Q_Goest]

Hi Jaap,
Thanks for the pictures and sketch. At least I can get a fair idea what you’re doing now. Without a detailed drawing though, no one will be able to really analyze this vessel very well, so the only thing I can do is point out what is important about the design in regards to pressure containment.

The ID of a 12”-4 thread is about 11.7”, so either the vessel ID isn’t 12” or the thread isn’t 12”-4. MAWP for a vessel per ASME code (section VIII, Div 1) is:

P = S E t / (R + .6 t)

Where P = Maximum Allowable Working Pressure (do not exceed pressure)
S = Allowable Stress
E = joint efficiency factor (= 1 for machined or seemless vessel)
t = wall thickness
R = inside radius

I can’t look up 7075 in the BPV code because you’ve not provided the temper or ASTM code it was purchased to, but I’d guess you’re talking about T6 temper in which case, and I’d guess the stress allowable is between 12 ksi and 15 ksi for this. If you can be more specific, I could look it up.

That puts the vessel rating up around 2600 psi, but there are lots of other factors.

The pipes/tubes you have coming in the top represent stress risers and they should be looked at. The size of the holes takes away from material resisting the hoop stress. This will lower the vessel's rating.

The holes in the middle of the screwed on heads would seem to prevent turning this into a pressure vessel, so how they are sealed, and stresses resisted at those points has to be addressed.

Regarding the thread itself, I wouldn’t use more than 4 threads to calculate the shear force on something like this, maybe only 3. The deformation in the parts won’t allow the other threads to contribute to the load. Without a whole lot more detail, I’d suggest keeping internal pressure below 1000 psi because of thread shear stress.

It’s unclear what the bolts are doing on this, and what cross sectional dimensions the end plates have. Without knowing a lot more about the dimensions on the end plates, the holes, and why the bolts are there, it's impossible to say what the pressure rating of the heads themselves are.

Note that almost all states in the US now require vessels that meet the description of the ASME BPV code be code stamped. The ID of this is over 6” and if pressure is over 15 psi (I think it is) then this vessel, per state law, needs to be a coded vessel. If anyone gets injured or killed by this vessel from overpressure, the lawyers will have a field day.

Have you considered doing a hydrostatic pressure test on this vessel to at least confirm what pressure it can withstand? That might be the easiest and most accurate thing you can do and would provide a degree of safety much greater than any free analysis you can obtain on a message board. Do a pressure test to 1.5 to 2 times the pressure you want to use the vessel for unless temperatures exceed ambient, in which case, get the material data and derate according to ASME codes for the material. That part is fairly simple.

 

[FredGarvin]

It's tough to tell from your print, so I'll ask for a couple more details. I need to make sure that this is what you are doing:

- Aluminum pressure vessel with pilot ID's on both ends.
- Steel threaded inserts inserted into the ends (How are they retained? Threads? Press fit?)
- End caps screwed into threaded inserts.

Does this represent what you are doing or is the vessel end cap being screwed directly into the aluminum housing?

 

[jaap de vries]

Pressure vessel cont'd

Goest I think you may have underestimated the strength of 7075 AL
This is what I got from Matweb
http://www.matweb.com/search/SpecificMaterial.asp?bassnum=NALIMEX06

Tensile Strength, Ultimate 360 - 540 MPa / 52200 - 78300 psi
Tensile Strength, Yield 260 - 470 MPa / 37700 - 68200 psi

This puts the pressure rating over 5000 (ASME BPV VIII div 1 UG-27)
. It needs to be about 3500. The holes on the top are NPT fittings going into swagelock plumbing.

Th vessel is seemless to.

The Threaded inserts are threaded straight into the vessel. Goest you are correct the inside diameter is more like 11.75.

There are two 8''OD 2.5''thick fused quartz windows clamped between two gaskets. there is a endplate that bolts into the threaded insert.

There is a steel reinforced, concrete filled cinderblock blast wall that surrounds the vessel.

I will do a hydrostatic test according to ASME BPV section 5. Any advise?

How do I calculate the threads load capability? That is my biggest issue.

Thanks for all y'alls help guys

Who can stamp this vessel for me?

 

[FredGarvin]

The shear of the internal thread is going to be, most likely the weak spot of the assembly. In that case, the maximum shear force can be calculated through the following:

Shear Area:
A_{s}=\pi*d_{min}*n\left[\frac{1}{2n}+\frac{1}{\sqrt{3}}\left(d_{min}-D_{2,max}\right)\right]

where

A_s= Shear Area
d_{min}= Minimum Major Diameter
n= Threads per Inch Engaged
D_{2,max}= Max Pitch Diameter of Internal Threads

The calculated shear force for the internal threads is then:

F_{max}=\frac{1}{2}S_t*A_s*L_e

where

S_t= Shear Strength of the Material
A_s= Shear Area from above
L_e= Length of Engagement

One note I would impress is that the design criteria is for threaded fasteners and the UN series threads (FED-STD-H28).
I can not comment on the applicability of this to such large threads or something that is not, technically,
a bolt or nut. It should get in the area you need to be though. With an ample safety factor you should have
your bases covered.

BTW...those numbers look good for 7075 Al. Just remember what I mentioned in regards to the temperature effects on the material properties IF you ever go to a higher temperature.

 

[jaap de vries]

Ok let's try this;

n = 4
St = 37700 psi
L = 2''
(Female Threads)
Effective pitch diameter = 12.0 in
Pitch, p = 0.25 (4 threads per inch)
Minimum major diameter, d = 12.16238 in
Max diameter, D = 11.83762 in
L = 2 in

Then from equation (1)

A = Pi * dmin * n * [1/2n + 1/Sqrt(3) * (d-D)]
A = 47.76 in^2

Then from equation (2)

Fmax = 1/2 * S * A * L

Fmax = 47.76 * S = 1.8 Mlbf

When I cannot use the full length of engagement (see Goest's comment) then:

L = 3/4

Fmax = 0.675 Mlbf

This last number would result in a pressure rating of about 5970 Psi.

Does this look OK?

 

[FredGarvin]

I don't have a good gut feel for threads that large. But it does look like you have it right. Just double check your units. I would then divide that final pressure by a generous factor of safety before you go any further.

I would also yield to Q for any other recommendations. Pressure vessels are his neck of the woods.

 

[Q_Goest]

Hi jaap. Regarding the strength of the material, per ASME code, you should be using the allowable stress, not the yield strength. I had a brief look through Part D (properties) but without having the ASTM number, it’s hopeless. Section D doesn’t use common terms such as 6061 or 7075 since these are too vague to pin down properties with. They use the ASTM number and various types/grades and UNS numbers to properly identify the specific type of material. Once the material properties are known, they apply a safety factor to them and say, “don’t exceed this stress level”, which they call the “allowable stress”.

However, I can find values for 6061-T6 in the piping code (ASME B31.3) which is around 12 ksi for allowable stress. The BPV values for 7075-T6, if you were to identify the material, would be very close to the 6061 values in the piping code, which is how I arrived at the values for stress allowable that I gave above.

In this case, you’re interested in allowable shear stress. Shear stress is about 2/3 of yield, so don’t get confused. See for example:
http://www.roymech.co.uk/Useful_Tables/Matter/shear_tensile.htm

So I’d suggest using about 8000 psi as an allowable shear stress. I’m not exactly sure how the Code handles shear stress on threads though, I’m not heavy into pressure vessel design and usually the shear stress on bolt threads is not the limiting factor, it's the tensile area.

All this gives you a very conservative pressure containing component of course, and an absolutely miserable value for pressure. However, that’s what the code would tell you to do.

As for who can code the vessel for you, that’s impossible. It has to be designed, built and tested by a National Board certified manufacturer.

If you’re intent on using the vessel, I’d suggest as a minimum, doing a hydrostatic test on it. Pressurize it with water to 2 times the pressure you intend to use the vessel for. Hold the pressure for at least 5 minutes, and I’d suggest raising the pressure in 10% increments till you get to this hold pressure. Once the pressure test is complete, disassemble it and verify dimensions haven’t changed (which would indicate yielding of the material). You shouldn’t have any leakage either of course. Either would be cause for rejection.

 

[jaap de vries]

There is a big difference between in 60 and 70 series in Al that's why we picked it out.
How much water should I use for a hydrostatic test and how dangerous is that?
Jaap

 

[jaap de vries]

"I’m not heavy into pressure vessel design and usually the shear stress on bolt threads is not the limiting factor, it's the tensile area."

I am not sure what you mean with this?

 

[Q_Goest]

Hi jaap,

There is a big difference between in 60 and 70 series in Al that's why we picked it out.

I just checked, and you're right. There's quite a difference, I thought they were closer. 7075 is almost twice the strength of 6061.

How much water should I use for a hydrostatic test and how dangerous is that?

You just need enough water to fill the vessel and any piping, then a small, high pressure hand pump or booster pump (Haskel booster) to get it up to the pressure you're looking for. It's a whole lot less dangerous than a pneumatic test. I personally wouldn't be too concerned about standing next to it, given the test was conducted properly. Worst case is something ruptures, but with only water inside, there's very little energy release. You get a crack, a little deformation, and that's it. ASME code prefers the hydraulic test over a pheumatic one and puts different factors in depending on the type of test.

I am not sure what you mean with this?

12-4 may seem like a 'course' thread, but it's not. In fact, it's a very fine thread. It's the equivalent of having 96 threads per inch on a 1/2" bolt. If you only consider 4 threads, which is a general rule of thumb, the shear stress area divided by the tensile stress area is tiny for such a thread. If this were a bolt, I would expect the limiting factor for your thead to be the thread itself instead of the tensile stress area. It's not a bolt of course, but there is still a load on it similar to a bolt except the axial stresses are compressive instead of tensile. You may want to consider doing an FEA on the threads.

Regardless, the biggest concerns for your vessel are the shear stress area of the thread, hoop stress in the cylinder and bending in the head. Bending on the head can also be calculated using either ASME BPV code calcs or you can use Roark's.

Remember you don't want to use yield strength, you need a safety factor of at least 1.5 on yield and 3.5 on ultimate. Of course, you'll need to find out the temper of the material used to figure out the strength of the material. That's another missing piece of information.

Another bit of missing information is how the pressure is going to be developed inside this and what the maximum pressure it could see is given a failure of other components that create this pressure. When considering what pressure you need to rate a vessel at, you need to consider the system it is going into. It would be nice to understand how you are safing this vessel too, do you have relief valves on it?

- At the very least, I'd suggest doing the hydrostat test on this, and because no stress analysis has been done, I'd suggest doing it to 2 times the intended MAWP (ie: relief valve set pressure or maximum possible pressure the vessel could be exposed to). If this vessel is going to be used more than about 1000 times, fatigue is an issue (especially for aluminum) and I'd suggest doing this test on a regular basis (or just tossing it out).
- A better option would be to have someone go over the design and calculate stress and THEN do the hydrostat test.
- The best option is to ditch the entire thing and have an ASME certified company make you a stamped vessel. They'll do the hydro test to certify it.

 

[jaap de vries]

The thing at a University is that we would like to build our own stuff. We do have a burst disk that can be set at a pre determined pressure.
The maximum pressure is created by a single combustion event of the gas inside. (Like a single piston engine stroke). There is no other system component that could cause the pressure to rise even more (which is good).
For the hydrostatic test is it ok to fill it with water and then to top it of with a unregulated air bottle through a needle valve? At the event of rupture the needle valve opening will immediately choke.
The material is 7075-T6
Thanks for all your help! Ever have a question about shock waves make sure to let me know!

 

[Q_Goest]

I can understand you'd like to build this yourself for economic and other reasons, but pressure vessels are governed by ASME code and almost all states have laws requiring such vessels to be coded.

If you have a burst disk set at 3500 psi, I'd suggest you do the hydrostatic test at 7000 psi. The highest pressure you can get in a cylinder is around 6000 psi, so you might be able to use that but I wouldn't recommend it. Putting any kind of gas into the system defeats the purpose of doing a hydrostatic test, regardless of how you might control it. See if you can get a manual hand pump. http://www.grainger.com/Grainger/wwg/productIndex.shtml?originalValue=pump+hand+&L2=Hand+Pumps&operator=prodIndexRefinementSearch&L1=Hydraulic" for as little as $264 that will go up to 10,000 psi. Don't know if it's good for water, but if you have to use oil, no big deal.

Another option is to have someone do the pressure test for you. Try calling one of these pressure test outfits:
http://www.thomasnet.com/nsearch.ht...:+Pressure&heading=84852201&navsec=prodsearch