Modelling an LED screen

[Mike Wright]

I've a relatively complex rig where I'm very close to my loading limits for the roof.  One of the key elements of the rig is a large circular LED screen suspended from a circular truss.

 

Currently I have the screen modelled as a series of point loads, one for every pick up on the screen header bars.  My though is that this in no way reflects how a screen actually behaves as it is an inherently rigid structure attached to one which deforms under load.

 

Could I produce a custom 'truss' which replicates the screen header bar but has a rigid cross section and the same weight/m as the screen itself.  This could then be 'attached' to the circular truss using multiple pliable truss cross points to give a more accurate transfer of forces (the screen header will be rigged using round slings from the truss)

 

Any thought would be gratefully received

 

 

[SCParker]

I've not done this on a curved truss, but when I do a LED video wall, I add a distributed load equal to the weight of the wall. Most of the walls I've been using hang from a pipe underhung off the truss with webbing every two feet so, and then to the header bars. Thus the distributed load vs. several single point loads.

 

Disclaimer, I'm not an engineer, nor play one on TV. When I get close to the point limits, I punt to the house riggers and/or an engineer to double check.

 

hth, Scott

[Mike Wright]

*similar disclaimer* I am also not an engineer

 

Here is a demonstration of my theory.

 

On the left is a truss object created to be the same weight and dimensions as a theoretical screen .  It has a rigid cross section because as near as damn it that is how an LED screen behaves.  This is coupled to the truss using a truss link with a rigid cross section (mimicking say a half coupler*).  All but the end two links are in compression and have failed.  In reality this would prevent the truss from deflecting as expected unless you were using a pliable method of supporting the screen (spanset etc).

 

On the right is a distributed load of the same weight showing the correct deflection of the truss.

 

So in theory:

 

• If I use a rigid connection between header bar and supporting truss, I prevent the truss from deflecting.  Which braceworks cannot model.
• If I use a soft connection between header bar and supporting truss the load will be almost exclusively applied at the outer most connections where the deflection of the truss and the screen are equal.

 

 

Now in answer to my own original question: I have spaced my mother grid and the hoists that support the circular truss to minimise the variance in deflection around the circumference. I have achieved this without the need to apply any chain shortening as without load cells this will be wildly inaccurate in practice.  This means that in theory none of the supports between screen header and circular truss should be wildly in compression (the screen here is modelled as a distributed load applied direct to the circular truss) :

 

 

 There are factors like for instance adding additional hoists which can cause this deflection to vary massively and in this case chain shortening is the only option to restore the desired flat support truss:

 

 

My conclusion has been that to most accurately match theory to practice, my truss deflection must be as uniform as possible under FEA like Braceworks.

 

Excuse the lengthy rambling post but (especially given recent events) I am taking a personal (and have been asked to take a corporate) interest in the rigging of LED screens and am interested in peoples opinions.

 

I'm happy to have holes picked in any of the above or to be proven completely wrong!

 

*Pro Burger for those on the other side of the pond

Edited May 21, 2019 by Mike Wright

[Mike Wright]

I was hoping to get more response to this!

 

Never mind.  I'm pleased to report that Braceworks reports compared extremely closely to real world readings from load cells on the 4 points to the roof and measurements of deflection on the circular truss.