I-beam connector?

[ned]

Its not a welding table, its a CNC plasma cutter, 6 foot x 10 foot. THe floor you see is to hold water, when the plasma is operating the steel will be immersed in water.

What I'm trying to do is make it flat.

To achieve this, I am laying the beams on top, so that the weight and the orientation of the top beams pulls the lower beams (lower beams are the table) onto the same plane.

It's a long way of making a flat table. I think. No-one has criticised the idea yet so I still think its a good one. I am open to criticism though.

The top beams will be removed after welding and I will end up with, hopefully, a very very flat, and very very square, large cnc table.

 

[Screwdriver]

Yes I get that but the beams you place on top are effectively an "inverted welding table".

Other than the inherent dimensional stability of the more massive beams, I am not seeing where the advantage is on having them on top of the item to be welded as opposed to supporting them from underneath. Where is the reference plane? Gravity alone will shift a big beam like those and having them pressing on top of the workpiece is already distorting the target which you therefore support from underneath.

Surely you are then relying on the workshop floor to become one reference plane with the addition of jigs and supports transferring that accuracy to the workpiece, then add a second reference to the top which presumably you have to set up again to become a plane. But at that stage you are relying on the dimensional accuracy of the heavy beams placed on top. You also run the risk of introducing some distortion into the workpiece merely from the effect of having that much weight on top. You then somehow have to measure the flatness of the important surface plane (the top of the workpiece) which is now inside a sandwich.

So what I am not understanding is how that setup is better than the regular method for producing a flat plane (i.e. a welding table) then clamping your workpiece to that.

 

[ned]

Awesome, thanks. I'm in work so will take me a little while to read and digest.

 

[ned]

Where is the reference plane?

I dont have a reference plane, the floor of the shed is uneven.

__________________________________________________________________________________


Gravity alone will shift a big beam like those and having them pressing on top of the workpiece is already distorting the target which you therefore support from underneath.

I have them supported from underneath with jacks

___________________________________________________________________________________


Surely you are then relying on the workshop floor to become one reference plane with the addition of jigs and supports transferring that accuracy to the workpiece

No the workshop floor is uneven. So I have been using bottle jacks to make the table level, a bottle jack in each corner


___________________________________________________________________________________


Then add a second reference to the top which presumably you have to set up again to become a plane. But at that stage you are relying on the dimensional accuracy of the heavy beams placed on top.

I am totally relying on the dimensional accuracy of the beams, yes. The process I see it goes like this:

1) make the clamps 2) level the table 3) when I am happy that it is level, tighten the clamps

The clamping force and the weight will resist warping when welding

4) remove the clamps and the beams.


________________________________________________________________________________


You also run the risk of introducing some distortion into the workpiece merely from the effect of having that much weight on top.

Yes I realised that, so I intoduced more jacks mid-way along the span last night


________________________________________________________________________________


You then somehow have to measure the flatness of the important surface plane (the top of the workpiece) which is now inside a sandwich.

Excellent observation. I think I will raise the upper beams away from the surface by means of a spacer plate, that will allow me to run wire to check for flatness....


_________________________________________________________________________________


So what I am not understanding is how that setup is better than the regular method for producing a flat plane (i.e. a welding table) then clamping your workpiece to that.

I dont have a welding table. I don't work in this field, I work in IT and dont have access to engineering workshop equipment.
One benefit is, that if this method works, anyone can do it with mimumum tools.


I really appreciated your reply, and will welcome any more comments, suggestions, criticisms, it's all to the good, thanks )

 

[henry Kadzielski]

Forgive me for possible stating the obvious, a plasma table doesn't need to be perfectly flat, within a few millimeters is more than good enough. The torch height control will look after the variations, without drama, providing you have one

 

[ned]

Heya Henry, you are of course correct.

To me this is a large lego or meccano set.

This problem is the type of problem my brain enjoys chewing over. I just want to get it as correct as is possible for me.

 

[awemawson]

In my experience, the sheets and plates that you cut will have (shall we say!) "variations in flatness" or more realistically will be warped all over the place (!)

So as Henry says, just make sure that your torch height control works !

 

[ned]

I appreciate that, I've seen that myself, but, it should be to a lesser degree than smaller stock. Those beams should be pretty straight.

I'll do all that I can, and then I'll be satisfied that I tried.

 

[awemawson]

Well while you are getting things flat can you sort out the road round The Ring of Kerry, last time I drove on it I'd swear it had been rolled with a hexagonal roller, but I was driving my late lamented Austin Healey 3000 with about 1/2" ground clearance

 

[ned]

I dont live there anymore, Kerry, but I miss it I have to say, jeezus its such a good looking place. Going to college along the coast every morning was a treat.
Thats a great looking car the 3000