On the prototype in this post I used 3-1/2" fiberglass batt insulation between the back of the mirror & the plastic back on the trough.

   For the final unit I may have a local contractor spray in polyisolene insulation.

   Also considering double pane glass on the front for a dead air gap & better insulation.

   I think that the parabolic style solar water heater can work well even in a cold climate if the collector tube is protected from ambient air.

Any thoughts on insulation or glass for front cover would be appreciated.

  On the final unit I want to put 4 troughs on a tracking unit tied to a thermostat & a plc. Which like you said can be rotated out of the sun if temperature gets to hot.

  I also think with a little planning the tracking unit can be made to rotate out of harms way in a storm to avoid hail damage.

   The tracker can also be made to rotate the troughs down at night to avoid snow accumulation in the winter.

   Lots of posabilities!  Still in the R&D stage & having lots of fun!

Even with the 2" aluminum pipe we used if you cut down the flow rate or just simply fill the tube with water & let it sit, it will boil water & create steam.

Could also be used as a solar still for alcohol!

http://home.germany.net/100-441770/amsi-model.html

   There are numerous ways of calculating & laying out parabolas! I have seen excel spread sheets ,layout with a T square and many others.

   I downloaded a small freeware computer program from one of the solar cooker sites that gives a quick width / depth / focal point calculation with X & Y coordinates in a text file output that can be imported onto excel or into autocad.

   Unfortunatley what I have found out is that none of the spreadsheet type output programs output a true acurate parabola!!!

   They are handy to get the rough width depth & focal point but from there I draw the parabola in autocad where I can maintain accuracy within a thousandth of an inch!

  Once I have a true parabola I draw the bows & end plates & cut a Steel template with My CNC Plasma Table & used the Templates to cut the pieces out of plywood with a jigsaw. I know I lost some accuracy cutting with the jigsaw on the prototype but all parts on the final version will be cut on the Plasma Table & should hold a tight tolerance.

   I learned how to draw a perfect parabola in autocad based on one of My most prized possesions, which is an old 1940s  drafting & sheet metal layout book.

   The book was written based on hand drafting tecniques but is a breeze to adapt to autocad using automatic guidelines & snap points.

   I use a resolution of a 1/16" of an inch to set all of the points along the parabola which for the proto type trough was 384 intersecting points & guidelines. That many guidelines makes your eyes about cross when drawing the final parabola by using the fitted curve drawing function in my cad program but it is the only accurate way I have found to generate a perfect parabola.

   The c4 in the user name is by accident only there seems to be 3 other brianc's on the net and they must all have similar interests to my own! Plain ole brianc is always taken times 3 about every web site I visit!

   As for the temperatures that can be achieved check out Edmund Scientifics web site they offer a 2 foot diameter perfect mirrored parabola that will hold a quarter of an inch focal point & achieve 2,500 deg. F temperature!

   That is a good thought on using the pipe as the pivot point. Very elegant and simple and well worth incorporating into my final design.

   I have my first tracking system designed using some very small but powerfull pv cells driving relays hooked up to dc gearmotors that has proven to work out very well in the prototype stage.

   The first working model should be done next week and will be installed on a dual axis parabolic dish solar cooker.

   I will post pictures of unit in the coming week.

   The great thing about the design is that the relays will handle 7 amps of current at 30vdc so you can drive some big torque dc gearmotors if necessary. From the input side the relays only require 1.2 milliamp current at 12v dc. The small PV cells I am using to drive the relays are epoxy encapsulated and produce 8v at 44 milliamps in full sunlight. They have been very reliable at tripping the relays.

Harder to make than the simple one-curve that is proven to work well, but the claim is that you get full sun through most of the day without tracking.