If you were making a fixed bed (moving gantry) machine, then I would suggest the following:

Since you only need the Z-travel to machine plates 'stood on end', you can leave the bed at the best position for cutting what you will cut most often and just clamp the plates on the end of the bed to machine them. If the bed ends just inside the X-axis travel, then you can move the gantry (just) off the end of the bed for machining the plates.

Here's an example in case that's not clear:
Click image for larger version.  Name: DSC08525_medium.jpg  Views: 1780  Size: 511.5 KB  ID: 8871

If you put a vice of some sort on the end of the bed then it would be quite efficient.

However, since the bed is moving you wont gain so much from doing this, maybe 100mm less Z-travel required. That's still a big gain for virtually no change in the design. You might even be able to get the full 220mm length to fit by allowing the plate to pass between the bed supports. Clearly that severely limits the X travel in that configuration, but you don't need much.

You really need some diagonals on the back of the frame. Don't leave it open like that as it will make the stiffness parallel to Y quite poor. Getting the gantry sides cast would be nice, for the right price, however if this option is unrealistic there's no need to use 40mm thick plate. A solid cross section is a waste of material when subjected to bending forces, so in this case a box section would be better. For example you could use four 20mm plates to create box-shaped gantry sides and the difference in strength between that and solid material would be small. If you can give the dimensions of the gantry sides then I can say roughly how big the difference would be. The reason for this effect is the material towards the center in a solid cross section is subjected to very little stress, so contributes very little to the overall stiffness.