-
1
- #1
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations TugboatEng on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
multi load column buckling analysis 7
- Thread starter ozPE
- Start date
- Thread starter
- #4
I ended up using the total load at the top and total length. It is a welded T section with axial and single axis moment. Flange and stem both plastic, no issue with local buckling. Small section too so no issue with over design with the above approach for this application.
Alex Tomanovich is a legend by the way.
I was searching the wrong key words. Stepped column did it.
Thank you both.
Alex Tomanovich is a legend by the way.
I was searching the wrong key words. Stepped column did it.
Thank you both.
ESPcomposites
Aerospace
This can also be done with 1D Elements:
Brian
Brian
Agen666,
Read OP's question and provide straight answer if you feel can help, that's engineer should do. If you only want to find something to "criticize", go Facebook or Twitter to relieve your extra/waste energy.
Also, compared with people that like to say "yeah, we got COMPUTER, the program can even spit out a thesis that is non-exist, it solves everything", conservatism is better, when safety of the end users is at our hand - an engineer's first duty is to provide shield to people safely.
Read OP's question and provide straight answer if you feel can help, that's engineer should do. If you only want to find something to "criticize", go Facebook or Twitter to relieve your extra/waste energy.
Also, compared with people that like to say "yeah, we got COMPUTER, the program can even spit out a thesis that is non-exist, it solves everything", conservatism is better, when safety of the end users is at our hand - an engineer's first duty is to provide shield to people safely.
The best method would be a matter of judgement for me.
1) Even with half of the load moved down to mid-height, you're still only getting about 4/3 the capacity of the end loaded version. For any situation that I gauge to be less beneficial than that, I'd probably just run it as end loaded to keep things simple.
2) As a beam-column, many codified design methods will be amplifying the applied moment based on the ratio of P/Pe. For this purpose, it is not clear to me whether or not the same benefit applies as would be the case for straight axial loading. So, for at least this part of the calculation, I'd be inclined to use Pe based on having all of the load applied at the top.
1) Even with half of the load moved down to mid-height, you're still only getting about 4/3 the capacity of the end loaded version. For any situation that I gauge to be less beneficial than that, I'd probably just run it as end loaded to keep things simple.
2) As a beam-column, many codified design methods will be amplifying the applied moment based on the ratio of P/Pe. For this purpose, it is not clear to me whether or not the same benefit applies as would be the case for straight axial loading. So, for at least this part of the calculation, I'd be inclined to use Pe based on having all of the load applied at the top.
ESPcomposites
Aerospace
Not to get too far off topic, but one should always couple a classical solution (even if conservative) with a computational solution. You can then make a better judgement about whether the computational results seem valid (and also look at deformed shapes if available). Using two independent computational methods can enhance confidence. However, in the end, we can also take a step back and realize that these are Eigenvalue solutions. If the column is short or intermediate, the solution will overpredict the capability. And one must also understand how to address imperfections (there was a recent thread about this).
The point being is that classical solutions can be just as misleading as computational solutions if you don't fundamentally understand the problem (that is usually the biggest source of error). That said, an argument can be made that it is easier to detect when a classical solution is being used incorrectly than ensuring a computational solution is used/coded correctly (especially as it becomes more complex).
Brian
The point being is that classical solutions can be just as misleading as computational solutions if you don't fundamentally understand the problem (that is usually the biggest source of error). That said, an argument can be made that it is easier to detect when a classical solution is being used incorrectly than ensuring a computational solution is used/coded correctly (especially as it becomes more complex).
Brian
I think the loading is two classic loads … 1) axial load (like any column) and 2) a load uniformly distributed along the column, (like a shear panel) reacted at the base.
I would expect at the allowable for 2) is higher than 1), thus applying the total load as 1) is conservative. Possibly 2) is similar to KootK's idea (applying the distributed load as a point load mid-column).
I would suggest running the numbers once to see that this is so, and to see how conservative this assumption is; then you'll know rather than believing what's on the internet.
another day in paradise, or is paradise one day closer ?
I would expect at the allowable for 2) is higher than 1), thus applying the total load as 1) is conservative. Possibly 2) is similar to KootK's idea (applying the distributed load as a point load mid-column).
I would suggest running the numbers once to see that this is so, and to see how conservative this assumption is; then you'll know rather than believing what's on the internet.
another day in paradise, or is paradise one day closer ?
Denial,
My question was a bait, in an effort to identify the problem using questionary form, similar to asking straight forward "why's that, how it looks?", because the problem is unclear to me.
It is impolite to hijack a person's thread, and divert the attention to his own liking/disliking. This thread is not mine, if you don't agree with my opinion/method, then provide your own that is sticking to the focus of the OP's question/concern.
Precise? Read the comments from KootK and ESPcomposites, that are precise to me, and everybody has learnt something, agree or not.
My question was a bait, in an effort to identify the problem using questionary form, similar to asking straight forward "why's that, how it looks?", because the problem is unclear to me.
It is impolite to hijack a person's thread, and divert the attention to his own liking/disliking. This thread is not mine, if you don't agree with my opinion/method, then provide your own that is sticking to the focus of the OP's question/concern.
Precise? Read the comments from KootK and ESPcomposites, that are precise to me, and everybody has learnt something, agree or not.
-
3
- #13
I believe that, for this forum to function effectively, forum members have to be able to graciously challenge one another with respect to the quality and relevance of their respective ideas. That creates lots of opportunities for conflict since no man, in his heart, is 100% able to separate an attack on his ideas from an attack on himself.
One of the site rules here is actually that one member shall not annoy another. Given that, it's a miracle that I still exist here since I'm surely one of the top two or three most annoying members to ever have graced this forum. I've even read a book called Crucial Conversations in an attempt to learn how to disagree with people here without upsetting them. The techniques have worked well but, still, I get into trouble regularly. If you watch for it, you'll notice that I now preface every damn thing with "in my opinion...". I also try hard to avoid absolute statements because those tend to incense folks who hold opposing views.
I've come to conclude that a little conflict is just the nature of the beast around here. It's our inexhaustible capacity for forgiveness that keeps things running smoothly much more so than our capacity for flawless tact.
One of the site rules here is actually that one member shall not annoy another. Given that, it's a miracle that I still exist here since I'm surely one of the top two or three most annoying members to ever have graced this forum. I've even read a book called Crucial Conversations in an attempt to learn how to disagree with people here without upsetting them. The techniques have worked well but, still, I get into trouble regularly. If you watch for it, you'll notice that I now preface every damn thing with "in my opinion...". I also try hard to avoid absolute statements because those tend to incense folks who hold opposing views.
I've come to conclude that a little conflict is just the nature of the beast around here. It's our inexhaustible capacity for forgiveness that keeps things running smoothly much more so than our capacity for flawless tact.
I believe I did that, your reaction to my help was unwarrented. I don't necessarily subscribe to conservatism is always better, like I noted the approach you proposed/noted could be extremely conservative depending on the situation (which in this case isn't fully outlined). Sure if it's one member design that needs this treatment it might not be an issue, but if there are 1000 similar members you're wasting your clients money and using materials unnecessarily. Every now and then we have to push ourselves to learn something new and potentially do better as engineers.
The reality is the point I was trying to convey to the OP is that they may already have the tools required at their disposal to do this rather than resorting to a hand method, but perhaps they don't realise it. If they don't have any software that can perform a compression buckling analysis there is a free option mastan2 that could be used if the OP was willing to learn more about this method.
As espcomposites noted, you need to follow your code to make the correction to a design capacity from the theoretical bucking load (allow for 2nd order effects, imperfections, residual stresses and that sort of thing). This is quite easy but not well understood depending on the code being employed (as it seems to often be added as an afterthought as an alternative method that isn't taught at all in universities based on my experiences).
And retired - I disagree that's not a stepped column. Whether the load varies with length or the moment of inertia (a stepped column in a more physical sense), the numerical methods are essentially the same - you just focus on whichever variable is changing with x.
phamENG said:
That's a semantic stretch in my opinion. When I hear "stepped column", I'm thinking of a column with a cross-sectional change somewhere along it's length. I'd be inclined to call this a prismatic column with an intermediate compression load.
- Status
Similar threads
- Question
- Question