Printing large slablike pieces in ABS

There may be something I've missed somewhere that covers this exact topic, but I've seen a lot of discussion over adhesion to the print bed (temperature, glue, hairspray, etc) but I'm wondering about a specific problem.

The attached FreeCad model is something I'm trying to print in ABS. It was designed to be printed laying on its back.

The problem is that when I print, the corners of the slab toward the edge of the heated bed inevitably "boat up" once I'm up to layer 30 or so. By the time I'm done, I have a significant amount of warping up from the bed.

I've measured heating on the surface of the glass, and it varies quite a bit from the thermistor on the bottom of the heated bed PCB. When I read 115C at the thermistor, I read 95C on the surface *in the center* and 75C on the front edge, even less in the corners. I used an IR thermometer and read from above, pointing down toward the glass.

This big variance in temperature corresponds to the spots where I observe the model pulling up from the bed, and leads me to conclude a couple of things:

1. The design of the heated bed PCB may be lacking. It seems you should be able to lay out the resistance heating elements more densely toward the edges so that you get something closer to even heating based on the current supplied. But I don't know enough about electrical design to say for sure, just my gut feeling.
2. Perhaps I need to take a different approach, such as printing vertically and enabling support material in slic3r?

Any suggestions are welcome!

I don't have FreeCAD handy right now (a screen grab would help), but ABS is going to warp. If you get it to stick to the bed, sometimes it even pulls layers apart. A heated chamber is the only universal cure. For individual parts you can try printing other orientations, or design in features that break up the layers. Or print the part in PLA if that material is acceptable.

Thanks for your reply - you pretty much confirmed what I suspected. I've been thinking about a tweaked heated bed design that provides more constant heating, though...

I've attached a screenshot and one picture of the final assembly.

I ended up printing it standing upright with support structures enabled. I had no warping on the support structure for this one but for the larger fork assembly I had warping on the support structure, but it didn't affect the integrity of the final print (some of the support structure shrunk up into upper layers of support structure.

So far I'm pretty pleased with my first design attempt - there are more things to tweak then I can share this on thingiverse but probably needs a wiki page to explain everything.