Hi to all,

I have one query.....

What is the difference between Defining Anchor or giving displacement 0,0,0,0,0,0 or defining Restraint as X,Y,Z, RX, RY,RZ at particular node.


With Regards
Thanks

No technical difference.

But what is the reason for getting different result....
and result difference is almost more than 12 %.

May be this one helpful for u.

Anchor is providing full fixity for three translations and rotations about the reference axis. that means no movement of the pipes.


But restraints it limits the free movement of the piping.


I think it is the difference we get a different result.

Please Expert If i give wrong correct me?

A possible reason :
Anchor X,Y,Z,RX,RY,RZ have a standard stiffness.
This stiffness give mathematical not the same restraint loads as the way displacement/rotations made zero.

Thanx to all ...

Ohliger's answer is quite satisfactory...

But ohliger as i know restraint of any type takes stiffness what we define in config file.
Basically that is the stiffness of support steels.

I did one exercise...
I define 1st disp 0,0,0,0,0,0 to a particular node and then define a restraint ANC to same node ...
I got same result ......

But when i define Restraint X, Y, Z , RX, RY, RZ at that node ...
i got different result what i was getting earlier.

that's confusing me....

So, to take the effect of complete anchor or fix point, which type of definition is appropriate ?

Adi,

Read any standard FEM text book on imposition of restraint. This will take you to the background theory.

Regards

Normally "ANC" must give the same results as X,Y,Z,RX,RY,RZ thats right.
Different restraint loads i cant understand.
Because ANC restraint all directions and rotations with standard stiffness.
X,Y,Z,RX,RY,RZ do the same i think.

I dont no whats wrong.

Restraint stiffness and Restraint force/moment have allways a depedence.
Anchor can theoreticel not have zero displacement.
Very,very smal, but not zero.
A Anchor and a very very big Force must give displacement !

Hi Anindya/Richard
Can you please clarify the above point. I have done the similar exercise & got identical result for ANC/ XYZRxRyRz / 000000.
Although caesar do not consider absolute Stiff ( stiffness 1.0E12) anything since it is not practical. I understand stiffness consideration will be very similar in all above consideration.
Adi Singh can you please pass forward your caesar file to my mail id ( habibur21@gmail.com )
for my review.

Please check if a gap is entered by chance in the restraints gap fields.

Hi,

Please check if you had mention any friction factor in your conf. file settings. Because when you put the support as X, Y, Z and having some friction factors then there are possibilities that your result may come different.

If possible please send me your CII file on my mail ID (princekanugo43305@gmail.com) I had done the same exersize and found no difference in the result.

Thanks a lot guys for ur advice...
actually i were doing some thing wrong..
By mistake friction has been considered while taking restraint X, Y, Z

After removing friction i am getting identical results..

Yes, you should get the same results.

Shr,

For any unsupported body the stiffness matrix is singular ( no unique displacement for a specified load set). Use of restraints to remove this singularity can be done in two ways:In one way, when you minimize the functional those with specified values do not contribute. The other method is where a very stiff spring is introduced to simulate this effect.

The value of the spring stiffness depends on number of factors, some of them are:

1) Avoiding ill conditioning of the matrix.

2) Truncation and round off errors.

As I have mentioned earlier, any standard FEM text book should be referenced to properly understand this subject.

Regards

Hi Anindya
Thanks for your response.
Yes I will follow your advice on the issue.


Habib

adi Singh,
With the definition of anchor or X,Y,Z at a node, all the possible movements at that node is already blocked. Then why friction comes in to the picture even though entered inadvertently?