Flat slab on pile caps

[Anonymous]

Hi there all,

I have modelled a basement slab on pile cap foundations. The thicker pile cap panels have been modelled separately and connected to the basement slab panel via rigid links. I have then modelled elastic column supports pinned in the x,y and elastic in the z but free to rotate in all directions. When I run the analysis, I am coming across two problems: 

1) At the interface between the pile cap and the the basement slab panel (where the rigid link is) there is a sudden change in moment from positive to negative which doesn't seem right. See below:

2) When I look at the support reactions, I am getting forces in the x and y directions which should not be the case as there is only vertical loads. See below:

Can anyone assist me with these problems?

Thanks

Tristan

[Rafacascudo]

1- When you set direction as " automatic" you have to be sure what are your panels Local directions as Moment selected will be according to these local panel directions. In your case it is better to select cartesian Global X direction and select Myy . Then you can be sure that the moments shown for all panels will be on the same direction.

2- Are the sum of reactions on X and Y direction Zero?

If this doesn´t solve your problem ,can you send the model?

[Anonymous]

Hi there Rafael,

Thanks for your response. 

1) I have made sure that my panel directions are all the same and match with the global cartesian system.

2) Yes they sum to zero, but still doesn't make sense for there to be horizontal forces when only applying vertical load.

See attached test model.

Kind Regards

Tristan 

[Rafacascudo]

1-Model you sent is different from the picture on your 1st post. Results for Myy look normal to me

2- Horizontal reactions appear because you have 2 levels of slabs . Structure, then, works like a frame where horizontal translation is restricted at the supports.

Of course , if you set the slabs at the same level horizontal reactions will disappear .

Also , due to the different levels ,horizontal forces are acting on both slabs and changing the values of all acting internal forces.

Below compare Myy with the picture above when the slabs are all at the same level

On the model you sent slab and pile caps have the same thickness

[Anonymous]

My apologies, 

I was not able to share original model.

The main slab was suppose to be TH400. When you set the slab to TH400 and re-analyse. You get the following:

Is this change in moment due to the frame action?

What would you recommend is the more accurate way of modelling this? All slabs at the same level?

Kind Regards

Tristan

[Rafacascudo · Accepted solution]

I would do it at the same level. With this thickness difference , The small 1500mm slabs act almost like a rigid block/fixed support for the 400mm slab . You can test changing their calculation model to rigid body and check that the moments differences on the 400mm are very small.

See moment diagram below

[Anonymous]

Great!

Thanks for all the help Rafael!

Kind Regards

Tristan

[Rafacascudo]

Forgot to say . Always remember to freeze the mesh after you think it is OK. It can save a lot of time during the analysis process

[ledinhchinh0409]

hi. @Rafacascudo 

Can you attach the model here? Thank you!

[Rafacascudo]

@ledinhchinh0409   ,  @Anonymous  model is attached on the 3rd message of this thread

[ledinhchinh0409]

hi. @Rafacascudo 

I want to see the model in message number 6. I don't understand, how to build models for different thickness slabs but in your same elevation. Thanks you very much!

[Rafacascudo]

model in post 6 is @Anonymous model . It is done with rigid links connecting the different levels panels . It is done in a way that the panels top surface is at the same level.

What I suggested is to model the panels to have its CGs at the same level.

[ledinhchinh0409]

Hi. @Rafacascudo 

Sorry for mentioning the old topic.

In "webinars" 05,  @Artur.Kosakowski  advised to use rigid links to build models with two floors of different levels. I want your answer. Thank you very much.