Modeling Nail Laminated Timber with plywood

[Milan Marojevic]

Hello everyone,

I am currently working on modeling in Abaqus Nail Laminated Timber (NLT) basement wall with plywood on both sides (exterior and interior), similar to a sandwich structure. I would appreciate some advice on how to proceed with the modeling process.

My main concern is regarding the nail connections between the lumbers. Should I model each individual nail connection, or is there a simpler approach? Modeling every single nail connection could be quite time-consuming, especially considering the large number of nail connections if I were to model just a 1-meter-wide section.

If it is necessary to model the nail connections, I would greatly appreciate any advice on how to approach this task efficiently and accurately.

Thank you in advance for your help and suggestions!

 

[SWComposites]

What is the purpose of the model?
Have you done a simple sandwich plate analysis using close form equations?
What are you going to use for material properties?

 

[Milan Marojevic]

I am MSc student and dealing with timber basement walls design. In this case i want to analyze NLT wall with plywoods on both sides connected with nails under lateral soil pressure. I have material properties of NLT and plywood. NLT panel is made of horizontally placed lumbers connected with nails…
I will also try to do analytical model, but for analyzing behaviour of the whole system i would like to use Abaqus for validation

 

[SWComposites]

What are you going to use the model to predict? Stiffness? Deflection? Strength? Something else?

The purpose drives the way the model needs to be constructed. And frankly modelling every nail seems absurd.

Also, FEMs don’t provide “validation”; test data does.

 

[Milan Marojevic]

I need to predict structural performance (deflection, shear forces, bending moments) and compare that with analytical model, thats why i said “validation”… however, real test will probably also be done in future to have real validation

 

[SWComposites]

If you are only going to predict displacements and internal forces, then there is no need to model the nails.

 

[Milan Marojevic]

Yes I need. Failure is more likely to occur at the contact between lumbers in NLT or NLT/plywood than in the wood… homogeneous model is not representative

 

[SWComposites]

But you did not say that you wanted to predict failure. You need to be precise about the purpose of the analysis.

Even to predict failure you don't need to model each nail. You need to be able to extract the shear (and maybe tensile) forces at the interface between the timbers and plywood. Then given a nail spacing, and the strength of the nailed joint (which needs to come from some test data as the failure mode is likely complex), you convert the shear force into a shear force on a single nail, and from that predict strength for that failure mode.

 

[Milan Marojevic]

As I said I want to find forces under soil pressure and to compare them with resistance forces/deflection… However, under some loads (depending on timber thickness) failure will occur for sure..

Thanks for reply. Do you think that I should make the cohesive behaviour between surfaces where nails should be placed or general contact or to connect with couple of nails and then to extract forces?

 

[SWComposites]

If it was me I wouldn't mess with cohesive elements (have done it, can be a real pain) and contact seems like overkill and possible convergence issues. Just model the plywood facesheets and timbers with with shell element (this assumes the timbers are spaced apart and not a solid mass between the plywood sheets) and extract either nodal forces as the interface or forces/stresses in the timber shell elements (this assumes there are several shell elements thru the depth of the timbers to get a reasonable shear distribution).

 

[Milan Marojevic]

Thanks !