Greenhouse Project 3
Here’s a pictorial series on the Greenhouse Project, #3. If you missed the beginning of this series, here’s Greenhouse Project #1, and Greenhouse Project #2.
I last left off with the foundation built, “feet” installed and bolted down, and the truss legs attached. Right after this, I realized that I had a problem. All of the truss leg mounting holes, where the cross bolt passes through the mounting feet, were drilled incorrectly at the factory.
The pic below shows a file of the same diameter as the bolt hole, and the curve of the leg. They should be perfectly aligned:
ALL of the truss legs (except just 1) were drilled wrong!
The (serious) problem with this is when the truss was fully assembled, the arch of the legs was pointing off by 20 degrees (each side). The left side of the greenhouse showed the arch 20 degrees towards the front of the building, whereas the right side of the greenhouse showed the arch angled approximately 20 degrees to the rear. Combined, the angle was off by 40 degrees! This would have caused a tremendous amount of stress on the entire assembled truss (mounting feet bolts, joints, screws and cross-members) if “forced” together. The only solution was to redrill everything.
Redrilling was going to be a major pain. There were 42 legs to “repair”, a job I didn’t want to do. I couldn’t simply redrill the holes wider either, not without making them very loose. To make matter worse, removing the truss legs from the mounting feet destroyed nearly all of the cross bolts (several broke in half, many stripped out).
I contacted the factory and explained the problem, along with sending several pictures of the issue. Once we got through the misunderstanding and misdirection, they agreed that either the truss legs needed to be replaced or redrilled.
Replacements were going to take a couple of weeks, so I opted to redrill them myself. This proved to be far more challenging then I realized at first. Curved, round pipe does not lend itself to being positioned correctly (or consistently, from pipe to pipe).
I eventually hit on an easy solution. First, I cut the pipes 24″, shortening each leg. This lopped off the bad mounting holes and lowered the overall peak height, something I opted to do.
I “processed” all the truss legs the same to ensure final assembly accuracy, or so I hoped.
Once cut, the sharp edges needed to be deburred. Helping hands here at work:
Next, I jury-rigged a “jig” to hold each pipe steady before locating and redrilling the (accurate) holes needed for the mounting feet.
The curve of the pipe could be laid flat against a wall, which I didn’t have that wasn’t already in use, so I hung up a piece of plywood for a temporary wall. The block was positioned in relationship to the drill press table height:
This proved to be very steady. Someone could hold the end of the pipe firmly against the wall while the new mounting holes were drilled. This kept the pipe “square” to the curve of the tube!
A block (2×6) was used to accurately space how far from the end of each truss leg the hole needed to be. It didn’t matter if it was 5 1/2″ or 4″ — they just all had to be exactly the same.
The machinist vise I have is pretty crappy, but steady enough for this kind of work. The pipe was slipped in between the jaws, up to the wooden block and then drilled straight through with a 3/8″ bit:
They all came out far more accurately then what the factory did. Which made me wonder if I could do this so easily myself, why couldn’t they?
Days later (tons and tons of rain here, slowed everything down for nearly two weeks), I learned that my cut / redrill process worked perfectly. The trusses lined up exactly as they should.
A quick test of several legs showed the holes were now all properly aligned with the curve.
Now on to truss assembly. The factory told me that I should assemble the entire truss on a flat surface. So everything gets laid out on the ground like this (curved truss legs are still missing):
The center support (cross-bracing) has to have clamps installed at the peak:
More hardware to install and put together:
Every slip-joint where the pipes connect must pre-drilled and a self-tapping screw installed. You could skip the pre-drill step, but it proved to be easier to drill a small hole first.
Then you drive in the self-tapping screw to keep the slip-joint together. Please NOTE if you do this yourself – The instructions were absent on this – location of the joint screw MATTERS when you assemble the purlin clamps. I didn’t know this. It gets in the way when you slip the aluminum clamps over the pipe if not drilled in the right position. I wish I knew this in advance.
Each truss, with cross-brace took anywhere from 10 minutes to 20 minutes to build — once we got the “system” down of laying out all the pipe, clamps, screws, brackets and legs.
I spent a lot of time making sure everything was “square”, with the joints properly aligned and so forth. I shouldn’t have bothered…. as you will see in the next blog entry.
We wound up loosening everything — the 24″ shorter legs made the truss width less (legs have an angle to them), and with tight clamps in place, we couldn’t erect the trusses properly without stressing everything too much, so we had to loosen all the cross-members.
This is just about half the trusses assembled, ready to be erected into place:
The first four trusses now finally in place. All the cross members are “loose” to allow the truss to flex in enough to accommodate for the 24″ shorter leg height and less width this created.
If the legs had been perfect right angles (90 degrees), then none of this loosening would have been required, but they’re not. Each truss leg is slightly angled in.
All the truss legs are now erected (next entry), and no ground posts were used (shown in Greenhouse 2, they turned out to not be needed, simplifying the assembly), but the higher purlins still need to be installed. You can see the lower purlin installed in the picture above, it’s the horizontal pipe attached to the truss legs. Each purlin will cross over about 4 truss legs. Longer purlin pipes would have been better (imo).
The lower row of purlins are in — but I found another factory “gotcha” — they didn’t supply me with enough purlins!
5 purlins (8′ long) only measure out to 39′ 1″ when assembled! Aaargh!!! So off to the hardware store and they didn’t have anything that would work. On a guess, I wandered around the lumber yard and looked at the cyclone fencing pipe and found exactly what I needed.
Despite the incessant rain, the greenhouse floor has yet to pool any water. This may mean the french drain is actually working!
Problems discovered so far:
Mounting feet were designed for 8″ or thicker foundation wall. Redrilling the mounting holes seemed to solve that, but it would have been very helpful if they had INSTRUCTIONS (listen up factory) regarding how thick the foundation wall needed to be to match the predrilled holes and not blow out the concrete (as shown in the Greenhouse 1 – first mounting foot).
The mounting feet were also poorly fabricated. The welded bracket was not always square to the foot. Where possible, I accommodated for this by simply rotating the entire foot slightly before drilling the concrete holes. It looks bad, but it worked better.
Truss legs were all factory drilled wrong. The solution was either replacement or redrill. Redrilling required shortening the legs no matter what (or also redrilling each mounting foot too — something I thought would complicate things even more). Without the right tools, I’d not have tried this.
Cross brace mounting bolts. Several of the nuts are the wrong size (20 – 30%), requiring a “socket switch” from 13 mm to 14 mm, slowing down assembly. No excused for that, not my fault, more poor factory control I guess.
Missing parts: Purlins were shorted, and so was the peak cap (didn’t receive any at all). During the redrilling requirement, I worked out a deal with the factory for cross-braces, and the missing peak parts (not shown yet). They cut me some serious slack on price and shipping.
Foundation: My design is 24″ centers due to high snow loading here. Each truss should be exactly 24″ center-to-center apart. The foundation was 40′ from outside edge to outside edge. The first mounting foot was mounted on the inside edge of the foundation, loosing several inches. The last mounting foot was also mounted on the inside edge of the foundation, loosing several more inches (foundation was 6″ thick). The reason the feet were mounted on the inside was because the sliding doors will be on the inside of the foundation (not the outside) due to snow buildup. I don’t want to shovel snow away to slide the doors open in winter.
The slightly shorter overall length should not be a problem, but it did mean that as I mounted each foot 24″ from it’s neighbor, when I got to the end, it wasn’t 24″ to the last foot (it was about 21″ as I recall). So I split the difference, wanting a truss there anyway for strength.
The mounting feet are several inches across (5″ I think, but it’s late at night, and I’m not going out there to check it now, got a bear playing hide and seek with me). The foundation needed to account for this by being about 10″ longer (40′ 10″). This would have allowed for an exact 24″ spacing on each foot (center-to-center).
As it is, it won’t make any difference. The polycarbonate will need to be shortened a little on the end of the building. And I’ve got one truss mounted at 21″ instead of 24″. Big deal. It’ll hold more snow down there.
Square: Cutting / redrilling solved the “hole square to the curve”, but round pipe, slipped together and arched over 26′ span is like a piece of wet spaghetti. I mean seriously — there is a LOT of movement (flex) in each joint, including across the top of the arch. Self-tapping screws keep the slip-joints together, and the purlins have aluminum clamp-on braces, but you must leave things loose enough so you can “square” everything up properly.
Each truss must be exactly 24″ on center, so when you lay the polycarbonate down (4′ x 12′ sheets), you can drill the panels in place properly. So the trusses must be individual leveled, and also adjusted to be exactly 24″ CTC from the neighboring truss. You do this by adjusting the purlin clamps (and the wet spaghetti likes to wiggle a lot). This is the step I’m at now, squaring and adjusting the clamps to get the building properly aligned so the polycarbonate can be correctly installed.
This is NOT a factory problem as far as I can see. It’s a building problem found in round pipe! There is a lot flex in the joints too. Longer pipe piece would have been nice. There are 11 pieces of pipe in a 26′ wide span (including cross-braces), which is a lot of joints. If you were to design your own pipe-frame greenhouse — use the longest pipe you can! But I’m no expert, it just seem to me that it would make more sense.
Coming up on the next Greenhouse entry will be the erected building, squaring things up, purlins installed and if I get that far, end walls built and sliding doors hung!