Revenge of the Catch-22
Per my last post on the topic of solar panels for the Mighty Intrepid/FWC Grandby, I had a decent planned setup with a 100W solar panel at each corner of the roof, each panel being held with a 20.5-inch long ABS adhesive mount fore and aft, for aero and dependability reasons. I figured that would be long enough to catch a couple of structural roof ribs under each one, and apply an abundance of adhesion to hold to the roof securely. The only real limitation was that I could come only so close to the roof edge because of tapering of the roof’s thickness.The panels would be held to the mounts with horizontal screws. The panels might be closer to the start of the taper than I liked, but it seemed promising.
As for using ground panels to add solar power, stowing a solar panel in slides mounted under the Grandby’s bed overhang has been done for awhile. After all, that platform is engineered to carry a heap of weight, being a 7″ or more vertical aluminum extrusion wrapped around the bed perimeter. Plywood forms the mattress platform, and a 20-pound panel hung under it should be no big deal, right? Given that the camper is now installed and that drilling holes from underneath is hit or miss because of its closeness to the truck cab’s roof, I thought about attaching a panel-carrying set of rails underneath, held by 3M VHB (Very High Bond) tape and supplemented toward each end with screws, since drilling there from underneath is not a problem. But what was the facing surface under this platform? There are a few different versions of VHB tape, each tailored for certain surfaces. I emailed Four Wheel Campers to ask.
What I promptly got back was a reply saying, “If you could please call us at ___-___-____, we can discuss a number of issues at the same time and wrap this up for you.” Say what? I’d expected an easy answer and a wish of good luck. So I called, and was immediately patched over to a speakerphone where their Service and Parts guy and the owner of FWC, Tom Hanagan, were on the blower. Their concern was my intention to use adhesive, and the fact that the 3/4″ plywood used for the bed platform is faced with Formica on the bottom to ward off moisture. There was some concern that the VHB-backed fully-loaded new panel slides might overpower the bond that the Formica has on the plywood. After all, it’s not 34 pounds of panels at rest, it’s 34 pounds of panels bouncing violently up and down. You don’t want to unintentionally delaminate the Formica. Tom suggested that I might get away with it as long as a number of screws toward the end were added. I hadn’t told him that my plan was to give each panel its own carrier beside the other, not doubled up in a single rack. That would help a great deal. Issue solved, in my mind.
Note that his concern here is not to tell the customer what to do or not do, but to try to advise on any planned mods so that the end result is a successful one. Virtually everything on the FWC camper is designed to do its job well – and no more. That would add weight, and more weight requires more strength, which also tends to add more weight that needs even more strength. Modifications tend to head over into territory where something may well become compromised, and Four Wheel highly values its street reputation. The last thing they want is for a customer to strap something on or nail it down in such a way as to create a problem with the camper itself, and then start whining about it in forums. I’ve actually read a complaint about how easily the FWC rooftop luggage rack tubing can be damaged by wiping one end against an overhanging tree branch. Seriously. The writer did not wonder where that force would transfer to had the rack been beefed up, though he did note that the roof itself had been miraculously left intact. The FWC is built for what it’s built for, and hanging a minibike on garage wall hooks screwed into the camper’s back wall isn’t going to work.
Before hanging up the phone, on a whim, I mentioned adding a 17-pound, 100W framed solar panel at each roof corner, using adhesive mounts. I mentioned my successful lift tests using the high-rate lift assist struts, and that each panel would be close to a lift strut, making roof loading a structural non-issue. That didn’t go over well, not at all. The issue was not weight, as I expected, but the nature of the roof itself, and the naughty term I’d used was “adhesive mounts”. The roof is a flat skeleton of rectangular tubing, with an aluminum skin overlaid. In spots, you can feel the lack of full support underneath by pushing down with your finger. Tom’s concern was the likelihood of any panel to lift or shake during travel, combined with the fact that the fatigue resistance of aluminum is pretty miserable compared to steel. If the mount is allowed to cause movement, the sheet will crack and leak at the flex site, requiring a tortuous skin removal process.
Mind you, I’m thinking about rigid strip mounts that are 20.5″ x 3.5″ in contact, and thinking that this will be a bummer if that aluminum roof sheet is just loosely draped over the top of the ribbing below. Tom was talking about how solar panel mounts should be screwed down to support ribs on the FWC, never adhered to a floppy aluminum skin. The problem now is that there’s no really accurate way to locate those ribs to screw into on a finished roof. You can get an approximation by feel, but when it comes down to drilling and screwing, it’s a gamble. Double bummer – must drill and use screws, but exactly where? Might be why the overpriced factory solar is so popular, and aftermarket roof installs are few. I asked about whether the top sheet is bonded in any way to the supports below, and it is: VHB tape. That’s very good because it prevents the sheet from moving on its own, and bad because if you do anything to hurt the sheet or cause it to vibrate, replacing it is a heroically arduous and expensive process. Thus sobered with the hard truth, I ended the call. Rib locations and obstructions would dictate how much of the roof’s real estate would be usable, and my plan to simply pop four 100W panels up at each corner was trash if I couldn’t find the physical structure to mount to.
But waitaminute! Given a strong enough adhesive bond above each support rib, couldn’t I lay VHB on top, stick the panel mounts down at each rib, and get much the same solidity as using a screw? Long mounts would probably only need to span a couple of support ribs to make a VHB and aluminum sandwich, though more would be much better. Solidly mounted at those multiple points, there would be no loading of unsupported sheet between them. It would all be on the ribs. All I’d have to ensure is that all panel mounts span at least two structural ribs, that the panel and mount make a stiff assembly, and that the resulting panel locations “work” without interfering with roof vents or the solar wire terminal, or run off the start of the taper near the roof’s edge. Lack of success here means cutting way back on that power and convenience thing, so it should be an interesting decision to make.
But first, time for a little touchy-feely on the camper’s roof. A quick grope showed that only ribs running front to back were prominent, and that the most significant of these straddled the roof’s dual vents and stayed some distance from the roof’s perimeter. A vague edge taper that looked like it began 6 inches or so from the edge was actually a full foot away. That one detail meant that I was dealing with a mounting area much smaller than I’d expected, the pinch between edge and vent making large sections of the roof unusable for solar panels. I began to see why FWC limited their mounted solar options to one roof panel, albeit a sizable one.
FWC’s current roof-mounted solar panel is a 160-watt unit supplied by Zamp. That’s a pretty big panel, but how big, where is it mounted, and with what type of mount? Is there a photo anywhere showing the FWC’s roof rib layout? A call to FWC for the latter yielded nothing, as did a Google search. What did turn up were a couple of photos showing panel placement and mounts. The panel was between the two roof vents, oriented crossways and centered left to right. It tended to be located a set distance from the front vent no matter how long the roof was, and covered over the solar power connector. That made sense to me, as those few who had left it unshrouded had warned of rain intrusion and terminal corrosion. The SAE plug used in making a solar connection does not offer the same protection as the tethered plastic cover. What was notable was that the panel’s leading and trailing edges were supported by a full-length, continuous aluminum mount going from side to side and of the same length as the panel, and that this panel and its extruded aluminum mounts appeared to stop about a foot or so from the roof’s sides – right where the roof taper begins.
Tracking down this panel showed it to be 58-3/8″ long, which just happened to stop it and its mounts right on top of the outermost full-height roof structure. And there was a reference to just three screws being used to fasten each mount to the roof, one at the central rib next to the side of the vents, and two at the full-height perimeter. Those are lathered with sealant. Not a lot of overkill on holding power for such a large component, but it apparently works just fine. Unlike the long mounts considered in previous posts and comments on this blog, FWC was not bothering to try to intercept anything with their mount but ribs spanning from front to back, with the side benefit of blocking off air from getting under the panel during highway travel. Theirs was a common “Z” mount, a simple zigzag profile of aluminum that bolts to the back of a panel frame, lifts it off the roof and provides a flange to bolt it down.
So, it seemed natural to try to adapt this proven mount style in an attempt to at least double the factory’s solar wattage. After all, I’d begun with a 600-watt target using lightweight semi-flexible solar panels, and when those threatened durability problems, weight constraints cut back conventional framed panels to a grudging 400 watts total. But that was before the roof surface I could use shrank to slim pickings. Where was I going to find room for an additional 240 watts, still be able to lift it with the roof, and keep it from fatiguing the aluminum skin?
Since I was stuck with a raging head cold, the only way to explore this was to lay out the Grandby’s roof to scale in drawing software, then start plopping scale solar panels down, using only long side-to-side extruded mounts that are certain to rest on the three or more principal ribs that I knew about. I began with a single 160-watt panel and tried to supplement it without any success, then laid down Renogy 100W mono-crystalline panels, since I already have two of them. I managed to get a total of three of them up there without too much difficulty, but that was it. 300 watts max.
I then toyed with Aleko 12V & 24V panels, which are available in a very wide range of wattages and sizes. They tend to be shorter and wider than Renogy’s panels, at least in the 100W size. 24-volt panels double the voltage but cut the amperage in half, so they don’t produce any more total power but do allow sizable solar arrays without needing heavy gauge wires to the controller. They’re also appropriate for ground panels, which typically have to send their power down long runs of wire that resist amperage, but not voltage. Unlike the Defiant, neither advantage applied in the case of my Grandby’s rooftop, and topped out my existing MPPT controller with just three 100W panels, so I kept going with more common 12V panels on mind, to allow future expansion.
Three 100W Aleckos fit in such a way as to open up a sizable gap between them, one big enough to pack in a 60W panel. In fact, if weight were not an issue, additional 20W and 10W panels could also be jammed on to total 390 watts. The 360-watt configuration seems livable for both weight and power, but the only nagging question is the serious suitability of VHB tape for the conditions that it will face on the roof. The temperatures can get high, and the actual working contact patches are quite small, which is why FWC leans toward screws. Since the rooftop solar connector must remain accessible, their Z-mount or the panel itself must also be removable for servicing. In their case, the entire panel and mount assembly can be unfastened and removed. In my case, I can use screws too, if I’m dead certain of proper drill locations. If not, then I would need to get around the need for accuracy with VHB tape, and the mounts themselves are best judged as unable be removed from the roof in any practical way without damaging the skin.
But the question remains about the usability of VHB tape in this application. Just how much holding power is typical? That depends on the adhesive chosen, and in this case the #4941 family may be as good as anything as an example. It typically will take 22 pounds of pull from one edge to peel it back off, and somewhere over 85 pounds per square inch to pull the bond apart in one shot. Shear, or pulling the bond sideways, can separate it at 70 pounds per square inch. The largest panel on the roof will weigh 20 pounds, and the manner of blocking road airflow underneath it will greatly limit lift. Given three contact points per mount at a measly 1 square inch apiece should more than take care of needs. Maximum sustained temperature to hold that bond is 194 degrees, which should be enough, considering that the aluminum panel mount will act as a heat sink, and any wind from travel movement will act to cool the joint. 3-day water immersion tests show no effect on bond strength.
So for the roof, VHB tape in place of screws is a possibility. What could possibly mess that up? Thermal expansion shouldn’t be much of an issue, since the roof and panel mounts are the same material, and the tape can handle sideways movement up to three times its .045″ thickness. Add in a little paranoia, and the potential trouble points finally appear. Being a welded assembly, any three roof ribs are not going to be in perfect alignment for flatness. The tape and its foam center layer can accommodate some unevenness, but the tape wants to see a hearty 15 pounds of shove per square inch during attachment for the best bond quality, and whichever rib is lowest may not attain that without two or three times that pressure. A serious misalignment of ribs can result either in a failed bond at the low rib, or at best a tall stretching of the foam as the gap encourages the foam to creep over time. This creep becomes permanent, and can set in at very low constant loads. 3M advises that static pulling loads be kept under 1 pound for each 4 square inches of tape in order to avoid this deformation. VHB tape is made to resist momentary high loads, not constant loads for long periods of time.
That foam layer is also slightly compressible, and potentially introduces the same problem it’s trying to prevent: aluminum sheet flex and fatigue. If a length of VHB tape is applied along the entire bottom of the long panel mount, any roof sheet stuck to it will be held motionless. That’s good. However, with the panel’s weight in place and bearing down on the three support ribs underneath, the tape foam actually bearing the load is quite small and may compress very slightly on bumpy roads, like washboard dirt trails. The surrounding foam stuck to the roof sheet does not compress because it has nothing to compress against. The potential result is that the aluminum sheet on either border of the load-bearing area may be forced to flex, if only a tiny amount. The amount is not as much a concern as the effect of long-term repetition on aluminum, which has little fatigue resistance.
Washboard roads are famous for encouraging spontaneous disassembly of fasteners, and in this case will present some unknown degree of potential for roof cracks from metal fatigue. The only way around this that I can see is to limit VHB tape application to being directly above the VHB tape already present below the roof. Not knowing precisely where the edges of that lower tape are makes this an unrealistic approach. Decreasing the VHB tape’s foam thickness and compression would also help, but this would inherently limit its ability to conform to rib misalignment, and harm overall bond strength. So the same inability to accurately locate drill points for screws also has a potential effect of the use of adhesive panel fasteners. This roof sheet flex of perhaps a few thousandths of an inch exists in theory, but whether the result will be a problem in practice is hard to say, particularly because the VHB foam underneath the sheet may take up half of the sheet flex with its own compression. Should I take the adhesive approach over screws, I’ll be banking that this potential will not flesh out into an actual problem over time.
The creep (or permanent foam stretch) is worse news for trying to suspend the panels underneath the bed overhang in storage, using only VHB tape. Suspending 9 pounds per rail requires a 36-inch run of 1″-wide tape at minimum (4 inches per pound), and that’s without any bouncing or vibration. Adding screws at the rail ends starts to seem like a pretty good idea, since sooner or later, the foam would stretch enough to visibly hang below the Formica facing. How sturdily the Formica is held to the plywood above is also an unanswerable question.
So, my take is that, given roof panel mounts long enough to span several roof ribs, adhesive tape is a workable approach. It’s a gamble in the log run however, both from the potential effects of vibration on the aluminum roof skin, and the basic integrity of the bonds after years of exposure. Once applied in this way, the panel mounts must be considered inseparable from the roof. Whether adhered or drilled for fasteners, the mounts will consist of thin aluminum L-channel extrusions, with the panels attached by self-tapping screws applied through the upper sections and into the panel frames. Which way I will swing on bonding vs drilling will wait until the proverbial last moment, probably depending on my confidence level in locating structural ribs underneath where each panel mount must go. Suspense!