Dear all,
Due to different approach and standards I read, many confussion & questions that build in my thoughts. That questions are specific in input of restrained with or without gaps and friction and the effect of friction in CAESAR II.The following approach will be as follows:

1. Standard pipe support dwg. shows a gap of 3mm in the guide and line stopper support, but normally such gaps are not input in the calculation model.As for the guide support, friction is not considered. When the gap more or less than 3mm is needed in analysis, actual gap shall be input in the calculation model. Only when gap for guide support is zero, friction shall be considered.

Questions:
1. Why only gap of more or less than 3mm shall be inputted in the calculation model for guide and line stop?
2. Every time I use Guide restained and without providing gap, does it mean caesar II can acknowledge that it is standard
and assumed that there is 3mm gap?

2. At any line stop and downward (+Y) or upward & downward (±Y) (Y) restrained support (with or without gap) friction shall be considered.

Question:
1. Why friction is considered in line stop and Y restrained with gap while guide with gaps did not?



Please share your idea!



Nald

There are in the forum many past items about this topic. I invite you to do a quick search on previous posts.

About your questions CAESAR II serves you to recreate the real world but it will only help you to do a model. Your scope should be to create this model as close as possible as reality. But this is not allways the case.

In CAESAR I will suggest you not to put friction on guides and to avoid gaps as possible, since what you want to do, is to understand the line movements, reaction forces at restraints and other variables that in the real world could be far away different.

¿Go to the field and try to find a single guide with a real zero gap? You might find this if your guide has an adjustable mechanism. Even if one knows that real guides with zero gaps are very scarse, when you model a line in CAESAR you want to have few variables in order to understand the system.

Santiago

Dear Santiago,
I've tried to search many times regarding related post to my previous query, but unfortunately I could not found...
Can you give me such specific post?

As your suggestion says "I will suggest you not to put friction on guides and to avoid gaps as possible".
My question is, For this specific type of restraint, when is the role of friction applicable?

My conclusion is like this, I will only apply friction whenever two surfaces are in contact. If I will provide gap (let say 3mm) in input file without providing friction in CAESAR II, then I found out in the displacement report in operating case was exceeded or equal to the gap i provided.This time i will consider friction since they are in contact.
Please correct me if I'm wrong..!

To Mr Dave Diehl And Mr Richard Ay please give a better explanation regarding my query?

Nald

A 3mm gap is a construction tolerance - you can't count on this, so don't model it. Gaps are for actual gaps you design into the system, say 12mm in a few places (not every support). Those you model.

There are many post on friction. Friction exists and for some systems it is important that friction is addressed. However, friction is not a constant effect, it varies over time. To address friction properly, you have to perform a range of analysis: no friction, low friction, high friction. Then you take the worst case and make sure everything still works.

Friction
You will never know exactly the effect of friction. You should study what happen to your system if you have friction and if you do not have. I usually use the friction multiplier to make this analysis. You company should write in the document of the project which kind of analysis is required. But of course you are an engineer and not only a Casar data enter.

Gap
You will be surprise to see how accurate could be the fabbrication of supports, but you will never will 100% sure that the 3 mm gap you required will be 2 or 4.
Some companies ask to use no gap.

Be careful, becuase many stress engineers believe that just by using gap and biliner restraint you create a linear system. If you have friction, this is not true, but this is another story.
When you built the load case, you can set the friction multiplier to create cases with or without friction.

Thank you so much with your kind reply..!

By read this thread it makes me a little bit confused here about gaps and friction.

About Gaps :

As mentioned by Richard Ay :
"A 3mm gap is a construction tolerance - you can't count on this, so don't model it. Gaps are for actual gaps you design into the system, say 12mm in a few places (not every support). Those you model."

I think i have to disagree with this coz there is a lot of real situation when you should model this gap, example is with U-Bolt, there is a Gripped U Bolt and Non Gripped U Bolt, usually U Bolt modelling using a guide and if it is a Non Gripped U Bolt then you must input the gap for it. And an analysis result with and without a gap can be very different especially near an equipment, altough it is a 3 mm gap. Ok i understand that it is a construction tolerance and who can guarantee that in the field the support will be gapped by 3 mm, bcoz it is a small number to get in reality and pipe fitter can missed it by 1 mm or 2 mm, but this is what we are payed for.....to make sure everything is ok and safe As analysis, altough sometimes it is very difficult.

Another thing is about friction.....this is a discussion that i never found a satisfied answer. Some of friends here have mentioned that "Friction cannot or never be your friend"...Friction is unprecditable....but in reality there is always a friction. we cannot ignore them (thats is my opinion), take an example if i have a pipe rest on T-Post support or on a Goal Post support, there must be a contact between these two steel, and there must be a friction and we are all know that it has a value of 0.3, but....if i have a big plant with many of this rest support should i apply of all this friction? well if i do that, then CAESAR II will experience a convergence, so with CAESAR II we become confused to apply a realistic condition, and as result our analysis never get near to reality condition.
Why i never found this problem before with Autopipe or Caepipe?
SO the problem with CAESAR II to apply this non linear condition is the convergence and non convergence problem. There is a way to change a friction coeffecient in the configurator to a lower value, but is it a conservative way to do? won't our analysis become more far away from reality condition? i hope others that have more experience can share their experience. i would like to hear how our senior here like Mr. Sam manik, Mr. Anindya, Mr. John Breen, face this problem. how you applied a friction problem in your analysis? i hope i can get satisfied answer to enlighten me.

No Answer...after 7 Month...hmm maybe all are busy....

Since they are all bussy, I will try to give you some thoughts.

It is true what it was said by the others before.

Your model will never be so accurate that you can guarantee that reflect 100% reality. Therefore you must compensate somehow this.
How?
Good question. Pending on your ability and experience you will understand what is really important and what it is not.

You mention that for U-bolts there is a gap for sure. Yes, but can you guarantee that the gap will be equal in the field?

In the same time, assuming that you succeded to have the gaps as in your calculation (with an extra cost of......). Good, but did you consider in your model the real support flexibility.

Regards,

For U-bolts: I mean equal on both sides of the pipe

Tengku,
From my site experience, I found that in most of the cases, gaps (construction tolerances) shown on the support drawing were NOT provided. This is because these gaps are construction tolerances and not gaps provided deliberately. Since construction site is not a mathematical model, it's very difficult to maintain exact gaps (unless they are significant).

As far as convergence is concern changing friction coefficient is not the only solution. I personally don't prefer to change the friction coefficient at once. Study first why you are getting non-convergence. Try to change non linear restraint to linear. Most of the time (not always) changing support configuration will solve it.

"if i have a big plant with many of this rest support should i apply of all this friction?" Yes. We need to understand that every plant has critical and non critical lines. Critical lines are computer analyzed. For non-critical lines manual inputs are provided to civil. In both these cases (critical and non critical lines) friction is present. But friction in big bore/critical lines is significant. Also, if you are working on a big system you can split it into two or more smaller systems that are easy to handle. Care should be taken while splitting systems.

Regarding friction, it is unpredictable. Reason is that you cannot expect nu=0.1 or 0.3 throughout plant life. Environmental conditions, dust etc. will alter friction. This situation is of concern with sensitive equipments. That's why Mr. Richard mentioned to perform a range of analysis with different friction conditions. As you cited, friction should never be ignored but one should handle friction with care. You can refer peng paper on friction for better understanding.

Dear Sir.
In my opinion, gaps in guide support:
To consider Pipe OD thermal expansion and construction tolerance (standard gap=2~3mm)
Our company use gap=2mm in guide support.

In my opinion, gaps in guide support:
To consider Pipe OD thermal expansion(like a uolt or clamp shoe guide type) and construction tolerance (standard gap=2~3mm)
Our company use gap=2mm in guide support.
Most of gaps are Construction gaps.