Got some spare time on your hands, Ian?
There has been lots written on this subject over the years as I'm certain you've read. My single (small 60w) panel is glued to the roof via mounts bought from Jaycar. These have quite large feet giving adequate grip to the roof- it is the strength of the roof that really worries me! There ain't much there, mate!!

There has been more talk in the media recently about accidents on the Bruce Highway that have been caused by flying panels coming loose and ...... .
PeterandMargaret have a good setup with what looks like flexible panels attached to their window shields on their Ocka but these get hot if no air gap is left between the two panels, reducing efficiency.

I'm waiting for the "ultimate" solution to come out - painted solar! Tests were done years ago in the USA, what came of them I don't know. Possibly another "heat" problem or perhaps the Asian solar panel companies put them the out of business.

When gluing apart from cleaning property. Give the adhesive time to dry.

 

In the window industry. Glass bonded into frameless panels needs a month for the adhesive to set because it is 30mm thick as it takes about a day per mm to set.

Warren-Pat, you bring up the exact point, the tear-out strength of the surface that the solar panels mount to. While self tappers may have a high tensile strength, there grip strength is dubious depending on the thickness of the material they screw into, and the "Tear out" strength of alloy cladding or thin fiberglass is poor. That is why a large glued joint is better in both the latter cases, spreading the load over a large surface.
As for time on my hands yes I have a bit, just can't put it to productive use.
Ian

This is a standard coolroom shape. Use it vertically with spacer on the rear of the panel. Screw it on back the front on the leading edge of the panel.

I'm sure there are better profiles, but it is a start & it's anodised.

The roof of my MH is relying solely on the glue to stay in place.
So are the walls.
Oh dear, what shall I do?
Cheers,
Peter

Peter,
Carry a few big rolls of Gaffer tape?
You'll be right mate, afterall the Ocka can't break the sound barrier, can it?

A number of people have mentioned the importance of the type of roof substrate to an effective bond, but equally important is the material the mounts themselves are constructed from. Quality mounts are made of ABS plastic which common adhesive/sealants readily bond to, but this isn't the case with some inferior products being sold on eBay.

I bought some very common mounts from Ebay, advertised as being ABS, but when they arrived they were made of a translucent waxy type of plastic that I suspect was polyethylene. As anyone who has tried to mend a polyethylene kayak or kids toy will know only too well, it is almost impossible to get anything to bond to polyethylene. So, I did an experiment of trying to glue two of them together with three common flexible adhesives used in the RV industry - Sika, silicon, and a polyurethane adhesive sealant. None of them would bond to the mounts and they could be easily pulled apart! The same would have happened if I'd attached them to my roof!

I wrote to the seller and asked what they were made of and what adhesive they advised, but by their reply it was clear they didn't have a clue, and cared little. So, I sent them a picture of my little experiment (same picture as below) and explained they were unsafe for the purpose advertised and recommended they withdraw them from sale, which they completely ignored, just offering me a discount! Unfortunately it seems many eBay sellers just jump on the bandwagon of whatever's trending popular at the moment, but have little knowledge about the design requirements and intended use of the products they are selling.

I left a comment on the seller's page warning others about the product and consigned mine to the bin where they belonged, and bought proper ABS mounts from Jaycar for about three times the price. I was happy that I'd dodged a potential bullet, but hate to think how many of these mounts are already in use and are an accident just waiting to happen.

I'm not suggesting the terrible accident recently was as a result of inferior quality mounts, we will only know the cause of that one was when the results of the investigation are released, but I just wanted to warn anyone buying mounts that they should test them with their intended adhesive first to make sure they are compatible and will form a strong bond.

Thanks all - I've found this to be an instructive thread.

I had no idea about the adhesive/polyethylene issue and I would never have thought of the downward wind deflection idea.

Polyethylene (and polypropylene) should generally be considered "un-glue able" for all intents and purposes. It is a bit like trying to glue a candle stick. There are some glues that claim some success, but personally I would never trust them.
Likewise, painting these plastics is problematic. It can be done by chemically etching the surface to create a bond as is done with car bumper bars, but not with 'normal' paints directly.
Both can be hot air welded, spin welded and ultrasonically welded to themselves though.
Cheers,
Peter

iana wrote:

---

F1 cars, what was proposed is a spoiler, mainly causing drag, and upping the fuel bill.

---

 I'm not suggesting 1000kg downforce.

It's more of a minor tweak to have a clean airflow over the top. It might even improve airflow in general.

Why I used long alloy L section so secure . My dwelling has 7 ply with rubber sealing roof . The holes that secure to panel is larger with big stainless washer to allow some expansion , movement or it breaks away . I used Sikoflex and stainless screws . On other surfaces Like alloy . I use an extra 2mm plate Sikoflexed and riveted to body . Then rivnut through the whole roof with Sikoflex .

Before I put the mounts with the panel on my roof, I gave both the mounts & the section of roof a good clean with wax remover.
Frames probably are a better way, but that adds more weight & those of us with single axles don't have much of that to spare.

Correct