Surface changes Day - Night

The following plots show the surface difference between maps made in daytime (on seven days between 20020827 and 20041111) and nighttime maps made on 20040909.

Maps made before this time are too noisy to show significant details (adjustments based on RxH3 maps and its reduction software began 20021104).

A conclusion from these plots is that surface changes between night and day appear to be very similar on different days. However the database is not very large.

In a Technical Report from 1995 (TR/001/03/FB), Fred Baas discusses some temperature measurements and holography maps made in July 1995. Among other things he compares temperare measurements of the four center beams (connecting the counterweight with the backing structure/surface through the cabin) and of the cone bars (connecting the counterweight with the backing structure going outside the cabin). Some rough model calculations relate the antenna surface deviations with the temperature difference between centerbeams and conebars and with the temperature gradient in the backing structure.

I have looked to new temperature measurements made in June 2004 during which month the surface was monitored by frequent 80 GHz maps.

It appears that the gradient in the backing structure is fairly constant - see e.g. here (this figure shows the change in temperature of the backing structure between opening the doors at 17h30 and midnight). However the backing structure at all times has temperature gradients, the inner parts being much warmer, presumably due to heat coming from the cabin. See e.g. here (look for 'Average temperatures during maps'), in particular in sector 10 there is a large gradient.

In this plot are plotted for June 2004, the average temperatures of the centerbeams (16 temperature sensors) and the conebars (12 temperature sensors). The centerbeams always have the higher temperatures. For 20040601 all sensors are plotted as grey lines, showing the large range of temperatures in the (partly isolated) centerbeams. Some of the sensors in the centerbeams even show an increase of temperature after opening roof and doors (with a maximum temparature at about 22h00).
The red lines show the temperature difference between centerbeams and conebars. At night the difference is about 6C; during the day the difference is much smaller. Therefore the dish is set for a situation where the temperature difference between both structures has a maximum. The dayly variation are almost identical, except maybe on a night which was particularly cold (e.g. 20040611).
The blue lines indicate when roof and doors are open (from ompobslog).
Some nights the telescope was partly closed due to weather or an ill observer. On some days the telescope was open because of observations of Venus. On 20040605 the doors were closed for a few hours 'to let the telescope cool down', whereas on 20040606 the doors were kept open. It appears that it is better to leave the doors open in such cases because on 20040606 the difference between centerbeams and conebars stabilizes earlier than on 20040605.

One conclusion from this plot might be that the cabin is too much isolated. Temperatures do not stabilize quickly and stay too(?) high. I wonder whether it ever has been considered to remove panels covering the cabin in order to let the cabin cool down easier, rather than isolate the centerbeams from the cabin (which apparently never worked well), and use big(?) fans to disperse the heat coming from cabin instruments. Maybe it is easier to look to the cause of antenna deformations rather than to try to understand every change in the surface from temperature measurements.