I had a Rostock, it worked well, but I think you may be missguided on it's benefits.
Like others, I moved away from it for a reason.
At any rate,
I fail to see the point of building a delta the way you are envisioning. Deltas are harder to enclose, and with the changes you plan on making, a delta is completely out of it's element. Deltas are meant for light weight, high speed, precise operation, and you plan on changing the light weight parts for heavy ones. A cartesian is a far better choice here as they use heft and mass to create stability, which allows speed and precision. You're coming at this from the wrong direction. Can it work, yes, it is a good idea, probably not. Even if what you want to do works, your printer is going to be pathetically slow.
Also, electronics, as mentioned, you need your electronics outside the chamber. You also want your motors outside. You will also want better bearings on your idler and make sure you have some high temp grease in your linear rail bearings or use Igus bearings. There's no avoiding some sort of bearing or bushings in the chamber, cartesian, Rostock or Kossel.
You say it will be minimal wear, I beg to differ. the weight you are adding is going to create a LOT of wear and tear, especially if you use 8mm rails. They are barely sufficient for a stock Rostock, some have said they can bend 1mm under use with plastic parts. Then you have the added heat, and chains... Chains wear and case wear far worse than belts. Belts fail usually becvause of stress and environment, chains, physically wear from friction. They also are not good at switching direction, they prefer to go one direction, and ideally maintain a consistent rpm. Switching directions, starting and stopping will beat the living hell out of your cogs and create a ton of wear. Ask bicycle riders about chain and sprocket wear. It only gets wors when you use smaller and smaller cogs.
You will also need an actual frame.
The vertical frame of the Rostock is the linear rails and they are insufficient even in stock form. The wood only offered minor stability, it wasn't really a structural component, and as such, the frame lost calibration almost any time it was bumped. This is why the 3DR uses an extrusion in between the rails.
Another problem is you said the rods will be under stretch and not push. The rods already push and pull, they work in unison. If you want it be a strictly pull system, you will need to rework the entire design. It's been done, but used a much more complex system and looked and worked nothing like a Rostock. I wish I could remember the name, but it was on the Google Deltabot forum months ago.
Lastly,
Don't run your chamber to run at 90c. ABS is still soft at that point, especially if it's still cooling from when you laid it down. It will sag. You would want a heated bed at 90-110c, and your chamber at say 70 or 80c temp for ABS, and a similar setup with PLA, 60c bed, and say a 45c chamber. Something like that. You want it to stay warm enough to not shrink, but cool enough that it coold enough to not sag.
Like others, I moved away from it for a reason.
At any rate,
I fail to see the point of building a delta the way you are envisioning. Deltas are harder to enclose, and with the changes you plan on making, a delta is completely out of it's element. Deltas are meant for light weight, high speed, precise operation, and you plan on changing the light weight parts for heavy ones. A cartesian is a far better choice here as they use heft and mass to create stability, which allows speed and precision. You're coming at this from the wrong direction. Can it work, yes, it is a good idea, probably not. Even if what you want to do works, your printer is going to be pathetically slow.
Also, electronics, as mentioned, you need your electronics outside the chamber. You also want your motors outside. You will also want better bearings on your idler and make sure you have some high temp grease in your linear rail bearings or use Igus bearings. There's no avoiding some sort of bearing or bushings in the chamber, cartesian, Rostock or Kossel.
You say it will be minimal wear, I beg to differ. the weight you are adding is going to create a LOT of wear and tear, especially if you use 8mm rails. They are barely sufficient for a stock Rostock, some have said they can bend 1mm under use with plastic parts. Then you have the added heat, and chains... Chains wear and case wear far worse than belts. Belts fail usually becvause of stress and environment, chains, physically wear from friction. They also are not good at switching direction, they prefer to go one direction, and ideally maintain a consistent rpm. Switching directions, starting and stopping will beat the living hell out of your cogs and create a ton of wear. Ask bicycle riders about chain and sprocket wear. It only gets wors when you use smaller and smaller cogs.
You will also need an actual frame.
The vertical frame of the Rostock is the linear rails and they are insufficient even in stock form. The wood only offered minor stability, it wasn't really a structural component, and as such, the frame lost calibration almost any time it was bumped. This is why the 3DR uses an extrusion in between the rails.
Another problem is you said the rods will be under stretch and not push. The rods already push and pull, they work in unison. If you want it be a strictly pull system, you will need to rework the entire design. It's been done, but used a much more complex system and looked and worked nothing like a Rostock. I wish I could remember the name, but it was on the Google Deltabot forum months ago.
Lastly,
Don't run your chamber to run at 90c. ABS is still soft at that point, especially if it's still cooling from when you laid it down. It will sag. You would want a heated bed at 90-110c, and your chamber at say 70 or 80c temp for ABS, and a similar setup with PLA, 60c bed, and say a 45c chamber. Something like that. You want it to stay warm enough to not shrink, but cool enough that it coold enough to not sag.