Setting the axial tilt of Earth at 90 degrees, results not as expected

[djlea1]

Because I can, it's interesting to see what happens if I change the Earth's axial tilt to 90 degrees. (Motion > Rotation > Obliquity (relative to orbital plane). This puts the earth spinning on its side and means that the poles have the Sun close to directly overhead 24 hours a day for several months in their summer. In this situation it turns out, averaged throughout the year, that the poles receive more solar heating than the equator which for quite a lot of the year would see the sun low in the sky. I was hoping to get a situation where there is ice around the equator and ice free poles (at least in the respective summer). I nearly got there with the earlier version (although the Earth got a bit too warm to form ice anywhere) but the latest update seems to stubbornly keep the poles colder than than equatorial regions.

Has anyone else played around with this and know of any settings which I might be able to play around with to get the results I'm looking for.

[djlea1]

After I wrote this I found some settings which make a difference. The main one is the surface heat capacity (in the "surface" section). If I lower this (to a tenth of the default value) it makes the surface temperatures much more responsive to the seasonal changes. At the north pole I get a range of +20 degC to -40 degC. In the northern summer the above 60 degrees N the temperature is greater than 20 degC, while at the same time in tropics it is about -30 degC!

I still can't get my ice band around the equator though. Any equatorial ice melts at the equinoxes.