[SScnc]

I'm trying to determine the thickness of steel plate required to support a machine that weighs about 1500lbs. Details are:

6 inch C-Channels 78 inches long

steel plate on top 48" x 78" x thicknes ? (shown is 1/4")

Machine width is 36" x 76" @ 1500lbs. (represented by the block of steel)


My main question is, does the simulation show real deformation or exaggerated ? Or, do you only go by the scale ? I assume exaggerated as it does this even when 2" thick plate is used.

Thanks, this is the first time we've done this !

[cygnus x-1]

You can select a deformation scale or true scale by changing the definition of the simulation result (for stress and deformation results). By default it uses a scaled deformation that is variable based on the actual deflection. I normally change it to true scale and then look at the maximum deflection value of the deformation result.

C|

[SScnc]

Thank you for your help and advice. When I set it to true and IPS, it doesn't show any deflection, as shown below.

Can you please tell me what settings I need to do in the simulation to be sure I'm setting the weight correctly so the deflection is accurate ?

I have the block set as "plain carbon steel" and it's weight shows as about 1500lbs.

Thanks !

[cygnus x-1]

Thank you for your help and advice. When I set it to true and IPS, it doesn't show any deflection, as shown below.

Can you please tell me what settings I need to do in the simulation to be sure I'm setting the weight correctly so the deflection is accurate ?

I have the block set as "plain carbon steel" and it's weight shows as about 1500lbs.

Thanks !

The question you ask is simple, but unfortunately the answer may be quite complex. In order to get accurate simulation results you need to setup the simulation so that it mimics what you have in real life as closely as possible. Seemingly small details can make the results completely wrong. I'm not even remotely an expert in simulation but I've messed with it enough to know that it can be very tricky. So I make no guarantees on the soundness of my advice on this. How's that for confidence boosting?


Ok, so first I have a few questions.

Are you trying to test the C channels or just the thin plate?
Are the C channels rigidly attached to the floor, or are they able to slide? (bolted, welded, etc.?)
Are the C channels rigidly attached to the thin plate, or can the plate slide? (bolted, welded, etc.?)
Does the thick plate that you're using to represent the weight of the machine, represent what the bottom of the machine is actually like? Meaning is the bottom of the machine a flat area the same size as the thick plate? Or does the machine have feet? If the machine does have feet, the placement of the feet on the thin plate will have a significant impact on the deflection.
Is the machine rigidly attached to the thin plate?


What version of Solidworks do you have, and are you using the full simulation package or just simulation express? If you have the full package you can run the simulation as an assembly and possible get more realistic results, but the setup is more complicated. Given that what (I think) you're trying to figure out is relatively simple, an assembly is probably not needed. I ask what version of SW you have because if it's 2013 or newer I can put a simple example together and post it. Though if you have a student version that may also be an extra complication, I'm not sure.


C|

[keebler303]

It seems a bit convoluted to use FEA to solve a simple algebra problem.
I would go about it this way:

Assume a span of 48"
Assume width of 76"
Simplify the width aspect by dividing the total load into the width. Each 1" of width needs to support 1500/76 = 20 pounds (about)

Assume all 20 lbs is in the center of the span (worst case for deflection). Now you have a classical problem of a simply supported beam with a point load. Your actual loading is not this but you know the deflection will be less if the load is better distributed across the span.

Max deflection formula: Structural Beam Bending Deflection Stress Equations / Calculation Supported on Both Ends Single Load at Center

deflection = W L^3/(E x I x 48)

W is 20 lbs load, L is 48 inch span, E is 29000000 for steel, I is the moment of inertia, 48 is a constant.

deflection = 20 x 48^3 / (29000000 x I x 48)

Moment of Inertia can be calculated from equations here: List of area moments of inertia - Wikipedia, the free encyclopedia

For a rectangular cross section you have I = b x h^3/12
b is 1" for the 1" width of plate we are considering.
So I = h^3/12

Now put it all together:

deflection = 20 x 48^3 / (29000000 x h^3/12 x 48)

deflection = 2,211,840/(116000000 x h^3)

plug in guess for h of .5" and solve for deflection

deflection = 0.153"

for 1" thick

deflection = 0.019"

with some algebra you can specify a max deflection and then back calculate the required thickness.

this is ignoring the weight of the plate, which will cause the plate to deflect a bit more but you get the idea.

[SScnc]

The question you ask is simple, but unfortunately the answer may be quite complex. In order to get accurate simulation results you need to setup the simulation so that it mimics what you have in real life as closely as possible. Seemingly small details can make the results completely wrong. I'm not even remotely an expert in simulation but I've messed with it enough to know that it can be very tricky. So I make no guarantees on the soundness of my advice on this. How's that for confidence boosting? LOL ! I'll take whatever advice you have and trust it !


Ok, so first I have a few questions.

Are you trying to test the C channels or just the thin plate?
The plate, do determine the thinnest we can use.
Are the C channels rigidly attached to the floor, or are they able to slide? (bolted, welded, etc.?)
These 6" channels are welded to 8" channels below them. The space between the 6" channels is fully occupied by other equipment keeping the addition of other structural members from being place there and the height has been specified so we can't add structural members on top of the 6" channels so must rely on the plate for support.
Are the C channels rigidly attached to the thin plate, or can the plate slide? (bolted, welded, etc.?)
The plate will be welded to the 6" channels.
Does the thick plate that you're using to represent the weight of the machine, Yes represent what the bottom of the machine is actually like? No Meaning is the bottom of the machine a flat area the same size as the thick plate? Or does the machine have feet? If the machine does have feet, the placement of the feet on the thin plate will have a significant impact on the deflection.
The machine actually has a rectangle c-channel sub-frame with the same footprint as the steel block I'm modeling that will set on the plate.
Is the machine rigidly attached to the thin plate?
The machine's sub-frame will be bolted to the plate.


What version of Solidworks do you have, and are you using the full simulation package or just simulation express? If you have the full package you can run the simulation as an assembly and possible get more realistic results, but the setup is more complicated. Given that what (I think) you're trying to figure out is relatively simple, an assembly is probably not needed. I ask what version of SW you have because if it's 2013 or newer I can put a simple example together and post it. Though if you have a student version that may also be an extra complication, I'm not sure.
It's 2014 full and have been trying to run the full simulation on the assembly.


C|

Thanks very much for your time and knowledge helping with this ! I can zip the files and post them if it will help but as you can see, they are super simple.

Basically we are just concerned with the plate flexing at the few inches on each side of the machine, between the sides of the machine base and the edge of the channel flange.

Let me know if I can provide any additional detals,

Thanks again !

[SScnc]

Wow Keebler, that's impressive. My dad was an engineer and that's the same kind of way I saw him solving problems. However, I've forgotten how to do so many things like this because I've grown so used to computers solving things. Thank you ! I'm going to study what you've done.

Also, the ppl I have to submit this to want to see graphics and expect me to use their SW !

Thanks !

[cygnus x-1]

Excellent analysis Keebler. Being able to simplify the problem and get at least an approximate answer is a great way to double check your simulation results to make sure they make sense. One of the first things I discovered about simulation is that it's very easy to make cool looking graphics that are completely wrong. :wee:

SScnc, go ahead and post your files if you like and I'll take a look. Since you're just looking for deflection of the plate you can model just the plate itself without the channels. The questions I asked above are still relevant though, since having the plate rigidly attached to something at the sides will reduce the deflection significantly, as opposed to letting it slide like the simple supported beam problem does.


C|

[SScnc]

"One of the first things I discovered about simulation is that it's very easy to make cool looking graphics that are completely wrong." LOL, YES ! I discovered that quickly too !

I'm attaching them as a zip. These aren't from the model but are basically identical and is what is pictured in my first posts.

Thanks, I appreciate your time !

Steve

[cygnus x-1]

Sorry it took me a bit to get to this.

Here is how I would simulate the plate deflection on its own. Since you're only looking for the deflection of the plate it's easier to set it up as a single part simulation.
The simulation results are not included because the zipped file exceeds the maximum size allowable for attached files. You should just be able to open the simulation and run it though.

I set it up so that one side of the plate is fixed and the other side is able to freely slide. In the actual application this won't be true, but this gives more of a worst case value for deflection of the plate. In reality some of the load will be translated into tension that will pull the channels inward and reduce the downward deflection of the plate. This may or may not be an issue and you'll have to decide whether or not it is.

Anyway, have a look and see what you think. Hopefully this helps in some way.


C|

[SScnc]

That's very kind of you to do that ! Thanks very much. I'll check it out and reply.

Thanks again !