I am doing revamp job wherein one more air fin cooler is adding to the existing two coolers. The system should be symmetrical.
All the bundles are sliding as it is mounted on slotted bolts. So I considered all the nozzle points as restraint with +y in vertical and North to south with guide gap equal to slotted dimension and East to west with guid gap as to slotted dimension. I have run the anlysis without considering friction. Is this procedure is acceptable. Please some one can suggest me what to do.
Thanks in Advance.

[ August 17, 2001: Message edited by: GSK ]

Wow,

No friction, please let me know if you need investors for your new invention! Why leave out friction? Its real and varies tremendously. Are you trying to under-report your loads to make somebody happy? Or are you trying to present as real a model result as possible?

As for gaps, how much control will the gaps be built with? Will the bundles fit precisely etc.


You should model and look at results as real as possible and perhaps do sensitivity studies.... What if I get no gap in this direction or that? What if mu is .6 or .35? Once you estblish what things overload your model then you can deal with them.

Remeber flex analysis is not a video game trying to achieve some mystical numbers by computerized fakery, but it attempts to evaluate the worth of a design before its built!

FT coolers are very fragile and it is VERY important to be certain that the pipe loadings are minimized. You should not model the boundary condition as one (single acting) restraint, and certainly not in the +Y direction. The pipe to cooler interface is going to be either a weld or a flanged connection, both of which will transfer moments.

I suggest a model/boundary condition setup similar to the "vessel" examples we have in the Applications Guide. I would model to the nozzle, connect to the cooler with a CNODE, and then continue the model down to the base plates. (The CNODE will give you loads in the restraint report, and if you want to implement WRC297, introduce some flexibility here.) At the base plates you can apply friction and model gaps to simulate your slotted bolt holes.

[ August 20, 2001: Message edited by: rich_ay ]