A-34 #1 after trials, Kubinka
The 45 mm shells were fired from the standard 45 mm tank gun, the 37 mm shells were fired from the 37 mm Bofors gun in a Vickers 6-ton, both from 100 meters. None of the shells penetrated, but one 45 mm shell, deviously fired into the underside of the turret, managed to rupture the welding seam, jam the turret, and cause some spalling on the inside. The shell did not penetrate. The engine, which was running the whole time, did not stall.
Another test was carried out on T-34s engine, but this one was testing a different kind of protection. Combat in Spain showed that bottles filled with gasoline were an effective weapon against tanks. Soviet testers made no attempts to spare their subjects the brutality of war.
This test is described in the book A-34. Birth of the Thirty-fourth by I. Zheltov and A. Makarov, p. 183-187. This time, A-34 #2 was the victim. In order to ensure that the gasoline would ignite, the engine deck was covered with burning hemp. The T-34's secret weapon against incendiary fluids, shutters on the air intakes controlled from inside the tank, were closed. 500 mL bottles filled with gasoline thrown at the left and right air intakes or the center of the engine deck were ineffective, the gasoline did not penetrate into the tank. The engine did not stall, but the fire on top of the engine deck did have a taxing effect on the cooling system. When the test was attempted with open shutters, the result was much more severe: the exhaust pipes began emitting black smoke after 20 seconds, and the engine stalled. When restarted with the fire still burning, it worked for another 15 seconds and stalled again. This time, even after the fire was put out, it would no longer start.
The front of the tank was also resistant to burning gasoline. When bottles were thrown at the front of the tank, only small amounts made it through the turret ring and driver's hatch.
Here's another, more thorough penetration test, described in this lecture by A. Makarov. This is the first cast T-34 turret using high hardness homogeneous armour, produced at the Mariupol factory in 1940 on ABTU's orders.
Sadly, he does not describe the test in great detail, only mentioning that while defects were discovered due to an unfamiliar manufacturing process, ABTU representatives agreed that the designers were on the right course. Suggestions were made to improve the manufacturing process, and a few months later, an attempt was made again.
Note the inclusion of an observation device housing (on early turrets it was attached separately). The rear was also reinforced, and trials showed an improvement. In total, the turret was hit 32 times with 37, 45, and 76 mm shells. The robustness of the turrets was satisfactory. The commission remarked that even when hit with high caliber shells, there were no cracks or spalling. The resistance of the turret's armour, 51-55 mm thick, was considered equivalent to the 45 mm rolled plates on the old turrets, but could be produced much faster. In order to penetrate, a 45 mm gun had to fire at basically point blank range with a blunt tipped shell, or from 100 meters with a pointed tipped shell.
Another cast addition to the design was to the front of the tank, a connecting piece between the upper and lower front plates. This design successfully resisted not only 45 mm, but also 76 mm AP shells.
Trials of the connecting piece as well as dummy "upper" and "lower" plates.
The lecture also includes a tactical diagram of the T-34 against a 76 mm gun:
The tank was completely invulnerable from the front, but vulnerable from the sides, unless presented at a very sharp angle. Against its own HE shells, the T-34 would also have a tough time. A hit on the turret ring in the rear part of the turret broke through both the roof of the fighting compartment and the turret bay floor and threw the turret off the hull (it was torn off the turret ring by a prior shot). A hit to the lower side also destroyed the floor of the overtrack hull and dealt enough damage to the suspension to stop the tank. However, there were no penetrations of the tank's 40-45 mm thick armour by HE shells. This was deemed satisfactory by the trials commission, but there were some places that the commission pointed out had room for improvement: the front of the turret, the joint between the top and bottom of the rear plates, and the lower side not protected by wheels.