Stress analysis on downhole components (oil and gas)

[caseymessick]

Hi all,

I'm trying to build a model to accurately as possible simulate the stresses seen in housings in downhole components due to pressure. However, we have known products that are good in downhole environments that I continue to show failing, by a large margin, when simulated in solidworks simulation.

I imagine that the reason is that external pressures apply a compressive load which gradually thicken the material, so they can withstand higher pressures. Essentially, the model isn't going to accurately capture the loading conditions and material responses.

Does anyone have any experience here that could help shed some light and provide some pointers? From what I understand, there are lookup tables that I could use to calculate the stresses that different types of housings would see due to pressure. But I can't find it. Or if anyone has some pointers to use in FEA that would be most welcome as well.

Thanks.

 

[swimfar]

Are you able to share an example component and the material that it's made out of? I wouldn't expect size changes due to hydrostatic compression to have any significant effect, unless it's a very compliant material. However, it is possible for hydrostatic stress to affect the failure limit. How much it is affected will depend on the magnitude of the hydrostatic stress and the other applied stresses.

 

[caseymessick]

316 SS. Cylindrical tube, closed on both ends, ~3.7" in diameter, about 0.5" wall thickness

 

[rb1957]

things that work in practice and fail in models point to bad models.

"the model isn't going to accurately capture the loading conditions and material responses" ... indicative of a bad model, but you're thinking along lines that try to answer the question.

maybe someone's modelled these things successfully. If it's that much trouble (and that much value) I doubt they'll tell you much for free.

if not you'll need to do your own testing. Investigate real components where the model indicates failure.

What FE code are you running ?

Please don't multi-post. You should Red Flag your post (in the mech forum) and ask for it to be deleted (before someone else does).


another day in paradise, or is paradise one day closer ?

 

[FEA way]

Are nonlinearities included in your analyses ? It’s usually crucial to account for nonlinear material behavior (plasticity in most cases) and geometric nonlinearity. This can have a huge impact on the accuracy of the results and it can be easily overlooked when performing analyses in CAD-embedded FEA modules like SolidWorks Simulation that you’ve mentioned.

 

[caseymessick]

Sorry @rb1957, I'm new to these forums, didn't know that multi-posting isn't acceptable. I thought that different people might be able to offer different insights.

I am working to build and correct my model and I know that my model isn't good. Which is why I wanted to reach out and see if there's anyone here who has experience modelling components in high-pressure environments. I have run lots of different calcs based on different tables, I would love to know if there is a way to correct my numerical simulation somehow to accurately capture and reflect my hand calcs. The reason being is that my hand calcs are good baselines for general shapes, however, specific geometry that deviates from standardized tables could be an issue. I would love to accurately capture that in an FEA simulation.

@FEA way, appreciate the insight. I am sure that my model is not capturing nonlinearities. Are you suggesting that collapse pressure type scenarios could lead to non-linearities? How can this specific non-linearity be modelled? Any recommendations?

 

[FEA way]

Yes, I would definitely advise you to include nonlinearities in these simulations. Of course, this will make it harder to verify the outputs with analytical calculations that are based on linear solid mechanics theory but will give you much more realistic results. You should include plastic material behavior but also geometric nonlinearity which allows the solver to account for buckling and large deformations.

 

[caseymessick]

@FEA way I see what you're saying now. Thanks for the tip.

 

[TGS4]

I'll repost my comment from the Mech forum here - for when that other thread gets nuked...

You say that your analysis shows the products "failing". By what criteria are you demonstrating that they are failing? What failure modes are you considering?

Before you get into going down the road of non-linearities, etc, you need to ask yourself my above questions.

 

[caseymessick]

@TGS4 My original model and my model moving forward was failure at first yield. For a cylindrical body under external pressure, that's when the inside surface first sees von mises over yield.

However, it seems like with that kind of definition for failure, these components are going to be pretty over-designed. But because we don't have much restriction on size at weight at the moment, and are just looking for confirmation, I think this is satisfactory for me for now as I build initial prototypes and for first to market designs.

We already have slotted in a 2.0 improvement in the future, when that comes I would like to improve my model, perhaps introduce an elastic-perfectly plastic material model, and define failure when a certain strain is met. Some work will need to be done regarding how much strain.

Appreciate your comment and any further insight you have

 

[TGS4]

There are standards that you could follow. API 17TR8 for starters. If you are doing Design By Analysis, I recommend following the rules in ASME Section VIII Division 3 (or 2).

EPP won't even cut it. You will need to evaluate a plethora of failure modes: plastic collapse, local failure, buckling, ratcheting, and fatigue. Different design margins for each failure mode. The approach you're describing is much much too simplistic.

 

[caseymessick]

@TGS4 I appreciate your input, especially pointing me to API 17TR8, that will be a very useful guide. ASME Section VIII however applies to pressure vessels, which is a burst condition rather than the collapse that I am looking at, there could be something useful but the loading conditions are different enough that there may be better more applicable material out there.

I definitely agree with you that my approach is too simplistic, however, we already have legacy products that have successfully performed in very similar loading conditions. Enough to give me the confidence to go ahead and get a prototype built and pressure tested. I think fast iteration and testing will much better serve our needs as this seems the quickest to market to save the millions in revenue we could be generating on sales. If I have the time, I will look deeper into this in parallel to the testing I will be conducting.

Thanks again for the recommendation and advice. Super helpful.

 

[TGS4]

My reference to VIII-2 (my area of expertise) is for Part 5 - Design-By-Analysis. There's MUCH more than just worrying about burst. The upcoming 2023 Edition rules for buckling are probably the best in any industry (contact me offline for a copy of those rules).

BTW, API 17TR8 refers to VIII-2 and VIII-3.

Good luck. Most industries go with simulation to iterate on designs, then follow-up with testing to save time-to-market. you just need to do the simulation properly and understand it more than simplistic criteria.

 

[caseymessick]

@TGS4, Gotcha. ASME is definitely a weakness in my career so far and something I want to improve now that I have just changed jobs and industries.

How can I reach you? I tried emailing the email listed on your profile (the @asme.org one) but it bounced.

 

[TGS4]

Sorry about the email mix-up. ASME has converted email addressed. I have updated it now.