Update of my project--or how I’ve been spending my time when not blogging

I’ve not written much about my day to day activities at the Solar Center, so before (hopefully) getting back to some posts of a more editorial (and perhaps more interesting) nature, I’ll update my work here.

My primary project has been working on a portable, collapsible version of a solar cooker, similar in concept, but different in construction, to the cookers they are currently making here.  The concept was introduced the first week I was here during the Solar Energy course.  Dr. Richard Komp, one of the instructors in the course and one of the integral components in the development of the Center as it is today, had a general concept sketched out and introduced it to course participants.  On parallel tracks, he and I designed our own versions of the concept.  I built my design, and am also helping the women here construct a prototype with the sketches supplied by Dr. Komp.  In addition, a student group from Cornell (NY) is here this week with two of their own conceptual prototypes. 

One thing that has been reinforced for me is that (in the immortal words of those salesmen on the train in the Music Man) “You’ve got to know the territory!”  From an engineering perspective, this means that you’ve got to know the customer and his/her needs.  However, in this case, the customer is a bit nebulous as there is no real customer with a checkbook, but the customer is what each of us imagined him/her to be (i.e., who would want a portable cooker?).  Unsurprisingly, the designs pursued by Cornell and those here were a bit divergent.  Cornell interpreted the requirements to be a cooker that would be delivered in compact form and assembled by the user (sort of an IKEA approach).  I was interpreting the requirement of “portable” in the sense of being not only easily assembled, but also easily broken down again for moving (sort of a Coleman camping equipment approach).  Another difference in approach (which is actually a bit funny when I think about it) is that I, in Nicaragua have taken the approach of extending the technology and using items that may not be available here, at least yet (Susan’s trip was my means to obtain items that are still exotic here, e.g., polyisocyanurate) , while the Cornell group, with all the resources at their fingertips in Ithaca, tried to build within the constraints of what is currently available in Nicaragua.  Both strategies, of course, have merit, and this difference was really a conscious choice instead of a lack of communication.

In discussing the options for merging designs this week, the students decided to break into smaller teams and pursue one of each:  a “kit” that would be assembled one time only, or at most a few times, and a collapsible portable that could be assembled/disassembled quickly and multiple time.  The idea is to take ideas/concepts from all the prototypes and merge them into something better.  Theoretically, this is a good idea, but we’ll see if in the very tight time frame (only about 5 days of total design/build), if there is much synthesis.  I’m curious to see if we’ll really see a merger of ideas or simply a resizing/refinement of one of the original concepts.

Another interesting aspect of the Cornell group is that they are being filmed for a short documentary, so there is usually someone with a camera and someone with a mike at most times.  It feels a bit like reality TV, but fortunately, nothing so dramatic (at least yet).  Hopefully I’ll be kept in the loop and be able to find out how the film turns out.

The film crew as the Cornell group opened their boxes to reveal their designs.

My Design

(If you’re not an engineer [for example, my kids], you might want to sign out and wait for my next, (hopefully) more interesting post lest your eyes glaze over!  I’m lifting this description from a report that I’m writing for their records here, so I doubt that this will keep anyone on the edge of his/her seat.)

The cooker design uses a simple, wooden frame with glued and nailed lap joints at the corners.  All wood in the frame is nominally 1” x 1.25” and is the standard green pine commonly available locally.  “Nominal” in Nicaragua really means “very approximate” as the wood delivered is all ripped by hand and planed by hand with no jointer or surface planer to give a tight tolerance on dimensions.

Photo showing the frame construction with temporary strut to ensure frame squareness.  The insert is a close-up of the lap joint.  Excess adhesive can be removed with a utility knife after drying.

The skin, both inside and outside, is entirely made of thin, aluminum printing laminate (which is readily available here as scrap), with the exception of the protective cap, which is made of stronger galvanized steel.  The printing laminate is attached only by adhesive with no nails or screws.  Construction adhesive is used on the outside and silicone on the inside, where temperatures will be higher.  This is accomplished by pressing parts together while the adhesive dries.  The galvanized cap is attached by adhesive and by screws, like the current design. 

A view of the laminate after fixing to the frame.  You can still read some of what was printed on the back of the laminate.

Pressure applied to the laminate and frame while adhesive is setting.  When available, I also used 5-gal bottles of purified water for weight.  This general method is also used for attaching the Mylar to the frame (discussed later)

To toughen the outside skin, the rigid foam insulation (polyisocyanurate) is bonded directly to the aluminum.  The insulation has a thickness of .75”, so there is nominally a .25” air gap between the reflective surface of the insulation and the printing laminate on the inside of the oven.   The exception to the air gap is on the oven base, where a sheet of .25” plywood is installed to support the weight of the pot.  This plywood is bonded both to the printing laminate and to the insulation.  In spite of the thin insulation, the R-value is about 5, equivalent to about twice the thickness of fiberglass.

(left)  A view of the polyisocyanurate being cut with the reflective side out.  (right)  The insulation installed in the frames showing the 1/4" gat between the insulation and the internal surface of the cooker.

Both the reflector and glazing use Mylar, something that is new here.  The reflector uses a .002” silvered Mylar that happened to be in-stock in the Solar Center from some previous group and the glazing is .005” Mylar that Susan delivered.  Attachment to the frame is the same for both types.  The Mylar is stretched and attached to a plywood base sheet, then silicone is applied to the frame, the frame is then set carefully on the Mylar and wiggled a bit to ensure good adhesion and finally weights are piled on the frame while the silicone cures.  After cure, the excess Mylar and adhesive are trimmed with a utility knife.  For the reflector, only one sheet is required, but the clear Mylar is applied to both sides to give double-glazing.

(left) A view of the reflective mylar being attached to the frame.  The Mylar is stapled to the plywood to keep it taut.  (right)  A view of the mylar reflector from the back after trimming.  Press-fit braces were glued into the fram to stretch the mylar and produce as flat a reflective surface as possible.  The galvanized sheet metal cap is applied on this side.

The sides, back and door are all permanently attached to the base with hinges:  standard 2” hinges for the sides and “sash hinges” for the back and door.  Sash hinges are extra long on one side which allows the back and door to be folded back onto itself (against the exterior of the base) in the collapsed mode.  These hinges also allow the door to open without the need to elevate the oven above the surface upon which it is sitting.  (Longer, strap hinges can also be used to accomplish the same thing as the sash hinges at about one-half the cost based on US prices)  The sides are attached so they simply fold inward onto the oven floor.  The reflector and window panels are also attached with standard hinges along the back edge.

Add caption

Figure 6  (left)  Standard hinges for the sides are shown on top and the door is already attached with sash hinges and folded under the base.  (right)  The sides are now attached and can fold inward.  The door is in open (but not folded) position.

After the hinges are all attached, there are two separate assemblies:  1)  the base with attached sides, door and back and 2)  the glazing-reflector assembly (plus the black bottom plate, which must be separate).  These two assemblies are connected with both dowel pins for positioning and hooks-eyes for secure latching.  All protrusions are positioned on edges of the panels to allow everything to fold flat in the collapsed position.  There are two dowels in the back of the sides that insert into matching holes in the back;  this positions the sides properly.  In addition, there are four dowels in the top edges that are inserted into matching holes in the window, one each on the sides near the door and two on the back near the corners.

A view of one of the pegs on the side with mating hole in the back.  Also seen are two different types of gaskets.

Positioning pegs in the back (2 of the 4 pins that fit into the glazing frame) and a close-up of a peg and one type of gasket material.

Before the hooks are attached, the gasket material is applied.  Like the dowels and hooks/eyes, the gasket material is always applied to an edge rather than a face so the cooker will fold as flat as possible.   For testing purposes, there were three different gasket materials tried:  1)  a 3/16” silicone sponge rubber that was attached with silicone, 2)  a tubular, flexible, silicone  gasket that was attached with silicone and 3)  a plastic V-strip gasket that was attached with both silicone an staples.  These are other products that may or may not be available in Nicaragua.  The gasket material is applied to all four edges of the sides, to the top edge of the back, to the front and back edges of the base and to the front edge of the window frame.

Front view of cooker with the door gasket shown.  All gaskets are located on an edge and all shown here are high temperature silicone foam.

With the gaskets applied, the hooks/eyes are attached to the window frame is attached to the edges of the back panel and the edges of the door panel.  The hooks to the back are only to hold the window/reflector assembly down because the dowel pins prevent the back from opening.  On the door, however, the hooks both hold the top parts down and hold the door closed.  So to open the door, the front hooks must be unfastened.

Side view showing front (left) and back (right) hooks.  The back hooks are heavier and longer because when the door is open, these are th only thing preveting the front of the glazing frame from lifting.

Finally, the reflector support/adjustment rod is sized and fabricated.  It is made so the angle of the reflector relative to the glass (i.e.,  Mylar) can be varied from about 60° to 120° and additionally, in the closed position, it does not protrude beyond the frame.  This is to avoid parts protruding beyond the envelope of the collapsed cooker.

Top shows the reflector support folded and the bottom shows the reflector supported in the open position.

As another experiment to try other technologies, a black Teflon sheet designed to line the bottom of ovens was purchased.  While expensive, it is lightweight, very compact and will not break or bend. 

With the components all described, the cooker can be assembled.  It takes less than one minute to either assemble or disassemble the cooker.

Photo of the finished cooker assembled.

View of the finished cooker in the collapsed position.

Reflections on the Design and Areas for Improvement

While the design went together reasonably well, nothing is ever perfect and there were a number of things that came up during construction or during testing that might be improved upon.  Additionally, there are a number of items that have both positive and negative aspects so trade-offs must be evaluated.

1)      Tolerances on wood parts—
Because a collapsible cooker generally requires a better fit than a standard cooker so that swinging or sliding components will reliably function, the tolerances on parts is more important.  The wood generally available, however, is ripped rather than surface planed, so thickness is not controlled very well.  Additionally, the wood received is green (not dried) and tends to warp after machining.  With current technologies available in Sabana Grande, there is really no solution for this problem, but some strategies to mitigate difficulties are:

a.       Let the wood dry for several days or weeks prior to  building cookers then hand-select pieces for straightness and consistency

b.      Use jigs/stop blocks when cutting pieces to ensure consistency

c.       Make frames slightly oversized with the intent of planning each side to a straight, square condition.  This requires a plane in good condition, some level of skill in using a plane and a good woodworking vice (the vice at the Solar Center is a metal-working vice).  Alternatively, plane each piece flat before assembly.  A relatively cheap tool that could achieve some of these results is a hand belt sander with coarse (60-80 grit) sandpaper.  Again, this requires some practice to successful use.

2)      Moisture in the wood and windows--
When first placed in the sun, moisture immediately started to condense on the upper (cooler) pane of Mylar.  As mentioned above, the wood is green and therefore naturally contains moisture that will be trapped once the Mylar is sealed to the frame.  The liquid drops may have a negative impact on the cooker’s performance.  A vent hole might alleviate the condensation problem by letting moisture escape more easily, but would also be subject to moisture entering during the rainy season and insects/dirt entering at any time.  Experimentation here might determine if the frame can be initially dried with a vent hole then permanently sealed with silicone.  There may also be materials that would block liquids, dirt and insects, but allow vapor to escape.

Photo showing some of the moisture trapped between the panes of Mylar.

3)      Limited adjustment positions with rod due to short length--
This is a tradeoff issue because the goal to keep the reflector support short enough to be within the envelope of the folded oven means that the increments of angle adjustment are not very fine.

4)      Corner sealing--
Corners where strips of gasket material meet tend to have holes, meaning that the oven is not completely sealed.  Care and precision of making parts can help this and perhaps one could wrap the gasket material around the corners.  The danger here is that the material at the corner becomes so compressed by the bend that it still fails to seal.  With square corners, this is often an issue.

5)      Durability of gaskets--
Another sealing issue is the durability of the various gasket materials.  This will have to be determined by use and experience.  It will also depend on how often the oven is designed to be assembled and collapsed.  If more durability is needed, the currently-used method of molding silicone to the gap might be better, however, it will be challenging to get all the seals made simultaneously because there are so many (12 different surfaces need a gasket with the current design).  They should all be done together because the gap at every location is a function of how cooker is fastened together.  Another consideration with silicone is that it doesn’t have as much compliance (“squishyness”) as some of the other materials.  If the wood should warp after manufacture, the silicone will stop sealing whereas a more compliant material may still be able to maintain a seal.

6)      Location of dowels--
For the prototype, the dowels were placed at approximately the center point of the edges.  However, when gasket materials were applied, they sometimes had to be trimmed, so it would be better to place the dowels closer to the outside surface of the cooker to give as much room as possible on the edge for the gasket.

7)      Gap control--
It is desired to maintain enough gap between mating parts such that there is room for the gasket material to compress.  If the surfaces are touching before the gasket is applied, the hinged part may not be able to reach its desired “closed” position.  When hinges are fitted to the various parts, it may be advisable to use a fixed spacer to maintain the desired compression on the gasket.  (This also relates to the issues of tolerance, squareness, etc., in number 1 above.)

8)      Spacing of Mylar glazing and protection of the lower Mylar sheet when collapsed--
Because this design used a standard-sized wood for all the frame pieces, including the glazing, the spacing of the two sheets of Mylar is 1 inch.  From a heat transfer perspective, this is probably a bit too wide as free convection currents may be present in the space, reducing its effectiveness as an insulating layer.  Additionally, the Mylar flush with the frame is more susceptible to damage from hitting other parts.  A possible solution here is to make the window frame to an optimum thickness for spacing the two sheets, then add a thin strip on the bottom surface so the bottom Mylar sheet is recessed slightly into the frame.

9)      Mylar gluing--
For the current cooker, the Mylar was on a surface and the frame placed on top of it followed by weights.  With this method, one cannot actually see how well the silicone has sealed and filled the gaps.  An alternate that may be worth trying is to put the frame on the ground (or table) with the silicone adhesive up.  Then lay the Mylar on the frame and stretch until taut.  Then a finger of small stick can be rubbed on the bonded surface and the silicone guided so it makes a 100% bond between the Mylar and the frame.  The long-term durability of the silicone-Mylar bond needs to be watched as well.

10)   Assembly with adhesive only--
The method of attaching the printing lamina and Mylar without fasteners appears to be satisfactory.  It has the advantage of fewer penetrations through the protective skin, so the wood should be protected better.  On the other hand, it is important to ensure that corners are well sealed and durability testing should be done to determine if the life of the adhesive is sufficient.  An additional negative is that the construction takes longer since one must wait longer for adhesive to cure instead of using nails or screws to hold pieces together while the curing takes place.

So that’s an overview of my main efforts, but there are also a few other things I’ve worked on, though not all fall under a “mechanical engineering” heading.   One has been working on supports for a sagging roof on the restaurant.  One strut has been installed (but it did not have to move the roof much), a second strut has been cut and a third still needs to be fit.  I also need to wait for a time when someone can remove some of the roof tiles so we can lift the eave to a level postion.

A second small project has been to install a couple of stiffeners in the roof of the Solar Center.  These are to stabilize the roof in strong winds (i.e., hurricanes) and were supposed to have been done during construction, but didn’t get done for whatever reason.  I certainly want to get this done before I leave, because the next hurricane season will be coming up in about 4 months.

I’ve also been able to “tag along” sometimes and observe photo-voltaic installations, some natural building techniques and some drip irrigation installations more as observer than contributor, but it’s nonetheless been educational.  A couple of opportunities to go into “teacher” mode have also arisen.  For the CELL group, I was asked to do a one-hour, non-mathematical overview of heat transfer.   None of these students were science majors, time constraints and level of background were both challenges.  Hopefully Emily can give me some real feedback on how it went when the two students on the trip from Northland return to campus with a report.  After the presentation, I was also asked to put the talk into written format, so I now have written a complete heat transfer course in about 10 pages—not. 

For the Cornell group this week I’m involved more or less as a mentor with one team and next week MIT will have a group here, but I don't think I'll be as involved directly with them.  While observing the student groups going through, I’ve also been trying to develop some sort of short-term study abroad that I might be able to offer some time at UWP.  I’ve got a lot of ideas sketched out, but the key is getting it to fit somewhere in the curriculum so it “counts” for something.  Even if it doesn’t “count,” this place would be a great experience for our students and very eye-opening for many of them as well.