How to model FRP with software

[JoeH78]

Hi,

I need to model the FRP laminates + FRP wrap elements for RC columns and beams with software for design and retrofitting purposes. Have any of You done such thing up to now, even if it is not how would you go with software or with manual hand calculation?

 

[SWComposites]

First, where are you getting your FRP material properties?

Calcs for FRP composite laminates can be done with FEA or hand calcs. If you are not familiar with FRP then you need to get a laminate theory textbook and study it. Then you will need to get research papers on using FRP for reinforcement.

However the interaction of FRP on RC civil structures is a very specialized field, with only a few companies expert in it. Its not something that you just throw into a FEM and push some buttons.

What is the purpose of your analysis?

 

[JoeH78]

First, where are you getting your FRP material properties?
What is the purpose of your analysis?

Thanks for commenting,

Material properties are directly from manufacturers I mainly stick to the e.g. BASF MasterBrace® LAM 100/1.2 CFS or SikaCarbodur, Sika Wrap

The purpose of analysis is to rehabilitate the existing building with CFRP laminates and wraps.

Calcs for FRP composite laminates can be done with FEA or hand calcs. If you are not familiar with FRP then you need to get a laminate theory textbook and study it. Then you will need to get research papers on using FRP for reinforcement.

yes I know and have full plethora of books and nomenclatures in my portrfolio but I think that area of expertise still not reached the matturity because one way or the other it needs to be sustained with large experimental data which I'm missing, so that.'s why I refere to that forum. Surely I will not push some buttons and say voila but using computers for such heavy calculations composition of stiffness matrices definitely can not be done with hand calculations, I don't even mention that these are comprised per time step for example if you do the non linear time history analysis, with hand calculation you can not stand a chance.

Rehabilitation is defined in ACI 440.2R-17 but it is not totally clear if you have mezzanine floor and if you can not apply totally the longitidual CFRP laminates due to the corbel obstruction.

My point of interest was: Let's say that column bending capacity is not enough(strong axis with Eq combionation is governing) and I would like to improve flexuaral capacity of that column, in that case say that I applied L/4 of column longitidual laminates + wraps ( as strir ups Sika wrap width=50mm@150mm). That should definitely will contriibute to bending but how to model that in numerical manner and calculate it?

 

[SWComposites]

Ok, good to hear you have some knowledge in the area (I see way too many questions from completely uninformed and/or inexperienced people on these forums asking about very complicated FE analyses).

I’ve been doing aerospace composites for a very long time, but only have some basic knowledge of composite reinforcement for RC structures. From what I understand, composite reinforcement can be used for two basic purposes: confinement to prevent rebar breakout (eg, column hoop wraps), and strengthening (eg longitudinal plies along a beam or column axis). Both are going to need test data to validate a design.

Sounds like you are looking into strengthening. For this the issue is designing the reinforcement to share load with the RC column. Which means there needs to be a sufficient bond between the materials. And the ends of the plies need to be designed to avoid onset of delamination failures.

For basic modelling, assuming you are using a FE code the models the RC columns with solid elements, and assuming the code has shell elements with laminated composite capability, then one approach would be to use shell elements on the surface of the solid column. The effect on stiffness will be relatively straightforward to sort out. The effect on strength is more complicated, as there are a bunch of potential failure modes to check.

And what do you mean by “L/4”?

 

[JoeH78]

I’ve been doing aerospace composites for a very long time, but only have some basic knowledge of composite reinforcement for RC structures. From what I understand, composite reinforcement can be used for two basic purposes: confinement to prevent rebar breakout (eg, column hoop wraps), and strengthening (eg longitudinal plies along a beam or column axis). Both are going to need test data to validate a design.

I see totally different area,

And what do you mean by “L/4”?

-> one fourth of column length in vertical direction, because in earthquake governed load combination max moments are observed at the column top and bottom places I was wondering and inspired by the image attached, which was originally excerpted from ACI Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete StructuresReported by ACI Committee 440 , that is beam, but same logic should be applicable to column IMHO.

For basic modelling, assuming you are using a FE code the models the RC columns with solid elements, and assuming the code has shell elements with laminated composite capability, then one approach would be to use shell elements on the surface of the solid column. The effect on stiffness will be relatively straightforward to sort out. The effect on strength is more complicated, as there are a bunch of potential failure modes to check.

That is the interesting one, normally we use the 2 node ( 12DOF) frame element for frame beam modelling I can almost assure you that %90 of this forums engineers have modeled columns / beams as frame element in practical life not solids, not even shell. Shell we use normally for shear walls, foundations and walls but I'll try to modell it as solid if I can figure out and as you said numerous failure criterias von Misses, Tresca etc..

How do you model in aerospace the bonding resins in that case? let's say the airplane wing is damaged and you want to apply composite on top of it wrap textile is appilicable I believe but how resin is modeled ?

Thanks in advance,

 

[SWComposites]

Yeah, I figured the civil str world used beam elements. With beam elements there are two possible approaches:
- use one beam element with combined properties. Composite laminates are not fundamentally different from reinforced concrete. Basically you would calculate the EA and EI for the concrete, rebar and laminate separately, then sum them. Then get the strain and curvature from the beam results, and from those calc the stresses in each material.
- use coincident beam elements, one for the RC, and one for the composite. Post-processing results might be a bit easier, though harder to define the model.

Now, neither will give you the interface stresses between the RC column and FRP. But to do that takes a very detailed 3D FEA, and to predict interface strength one really needs to use fracture mechanics not stress based failure criteria. Extremely complicated. Thats why test data is needed.

For aircraft, composites are almost never used to repair metal parts (only exception I know if is the Aussie RAF used boron/epoxy repairs on F111 wings. But they tested the heck out of it.

We typically don’t model the resin separately (unless in some research projects); we test representative joint configurations to get specific design properties.

Is this a research project, or actual design activity?

 

[SWComposites]

Are you familiar with this:

https://www.concrete.org/store/productdetail.aspx?ItemID=440223&Language=English&Units=US_AND_METRIC

 

[Brad805]

You need software that will let you define the stress strain curve for the fiber. That is not your run of the mill software.

 

[SWComposites]

The stress-strain response of carbon fiber composites is essentially linear to failure, so not that hard to model.

 

[Brad805]

For you, yes, I agree. Most structural engineering software is geared towards steel and concrete where the failure is notable.

 

[JoeH78]

Thank you for shedding light on that topic

Are you familiar with this:

https://www.concrete.org/store/productdetail.aspx?...

Of course I'm familiar with that and it is the design activity

You need software that will let you define the stress strain curve for the fiber. That is not your run of the mill software.

how about SAp 20000, do you think that can handle it? Definition of iso tropic, anisotropic, time dependent behaviour is possible in that.

The stress-strain response of carbon fiber composites is essentially linear to failure, so not that hard to model.

Very useful pieces of info, it would be better of me to try to model and design with respect to that, at least to see the inital results.
If I define iso-tropic material which will reflect the FRP mech. properties
do you think that I have to use max tension stress and max elongation of FRP Element, BASF say for laminate E=165.000N/mm2 and fu = 3000N/mm2, elongation = %1.5 because that software allows me to do so,
the thing that I'm wondering how to model that is in case of compression, AFAIK FRP is not totally affirmed for compression it may heavily depend on debonding / delamination from surface.

 

[SWComposites]

FRP laminates are orthotropic, not isotropic, except for certain quasi-isotropic layups; I doubt you are using those. All civil str reinforcements I have seen have either axial or hoop direction plies (fibers) or both.

I would need to know the BASF laminate layup.

While tensile elongation can get to 1.5%, thats for a “perfect” layup. I would design to something much lower than that, probably 1% or less.

Compression strength can get to 0.8-1.0% elongation (strain), but value is highly dependent on layup, laminate quality, and assumes no premature delamination or debonding.

If this is for a real design, then there is going to need to be test data from somewhere to validate it. A FEM is not sufficient.

 

[Ingenuity]

Probably about 95% of FRP using in concrete related structural strengthening applications are unidirectional carbon layups. Most use wet layups, and a minority use pultruded strips.

In my experience - dates back to 1998 with FRP's - for such structures, FEM software is not commonly used. Most FRP manufacturers simply give the EOR maximum useable tensile capacity (based upon statistical analysis of actual test values), useable strains and bond capacity, and using ACI 440 design documents a design (flexural, confinement, blast etc.) is undertaken typically using using spreadsheets, apps. or MathCAD etc. to reduce some of the number crunching/automate the design.

I have designed and applied systems by MITSUBISHI, FYFE, BASF and SIKA - all of them used simply tools to design such strengthening systems.

 

[hardbutmild]

There is a software for strengthening of RC buildings, they have FRP wrappings with all the details built into the software. Since I do not do enough RC strengthening projects (and I'm poor) I have never tried it, but I have heard good things about it. It is called seismostruct by seismosoft (they have a few other software so look it up).

 

[Brad805]

If you are working towards a design that can be implemented, you will need to specify a product someone can buy. I would start by figuring out which products you can get and see what tech support they offer. Sika has a simple software program to help with some of this and I am sure others do as well. I am not sure what your goal is with this model.