On paper, this can be done by a piece of cable and a couple of pulleys. But I've never seen such setup in earlier published texts. Obviously this technique is only applicable for monotonic loading and small scale specimens. It might be possible to extend this technique to cyclic loading if you use two separate UTMs.

If your specimen is very small (less than 30 or 40 cm) you might be able to just rotate it by 90 degrees and place it between UTM's grips.

Dear Mr. Amin

Thanks for your valuable answer. In fact I expected some response from experts like you. Please see the attached figure where I drew a Shear wall in a possible support system. In fact such a figure is drawn to to test a shear wall in an UTM, giving compression only.

Suppose the length of the wall is 3' and width is 1.5'.

Can you express your opinion about it? also what may be the possible dimension of the support as well as reinforcement?

It's seems you gonna use the second technique (90 degrees rotation). The setup is a little weird but sounds sensible. You just need to consider a couple of things. First, used support should be stiff enough such that measured downward displacement can be attributed mainly to the concrete panel itself, rather than rigid body motion of the whole panel. Secondly, care should be paid to use a reliable connection between concrete panel and the proposed support. Apparently, you are going to use a concrete support. If it is, obviously enough development length should be provided for all longitudinal panel reinforcements. That is, your support should have enough depth, probably about 20 cm or more. So support depth (measured horizontally in your attached pic), should be selected according to stiffness or development length specifications, whichever is greater.

Besides, don't forget to provide some lateral braces both at the top of the support and at the free end of the panel. It is highly recommended to use some LVDTs both vertically and horizontally at the free end of the panel and at the top of the support. In other words, don't trust only on the built-in LVDT of the UTM.

About reinforcement details, it depends upon the intended aim of the experimental program. For example, if you want to resemble special concrete shear wall (per ACI 318, for example), you need to select some code-based reinforcement. In any case, I think you supposed to carry out some numerical analysis before the tests. In this way you'll be able to select the appropriate connections and supports and you'll also have a general idea about failure mode of the specimens.

Good luck in your experimental program

Thanks so much for your suggestion.