My next big repair job just walked into my lap. This is a finishing roller at a roof truss plant I do occasional work for. Assembled trusses a pulled in through 2 rollers that press the connector plates in completely. Basically anything you push in comes out 1.5” thick.
They called me today saying the bottom roller is cracking. So somehow I’ve gotta gind out as much of the cracking as I can and re-weld. Luckily the end plate is set into the roller quite a ways giving me plenty room to stack beads so Im thinking I don’t need to completely grind off the old weld. They were hoping I could do it all in place without disassembly but I can’t see that happening.
I don’t have air arc equipment, I’m guessing that would be useful. To reweld I figure I’ll use pulsed mig, maxing out my PowerMig 300 and stack a bunch of passes to make sure it’s solidly repaired. Either that or 3/16 7018 rods but I’m guessing the pulse mig would be a lot faster.
Any ideas would be much appreciated! I’m loving challenges like this, but I’m 24 so Im not the most experienced welder around. Love learning from you guys.
Dodgy. I think you really need to ascertain the composition of the roller core and the outer face. It could well be the face is an alloyed steel that is not friendly to welding. The crack could be the result of improper pre- and post- weld treatment.
One thing is for sure, once I knew the material and treatment, I'd grind a V groove to remove the old repair and get into fresh clean metal. I would choose to lay a thinner root with even a 3/32 or 1/16 wire (stick or TIG) then fill from there. May not be necessary, but I would want to drive new wire in deep and fully, under the correct temps. Or it will likely just crack again, and again.
Like CJ stated, I'd bet that is some type of alloy. Can you take that drive chain loose so you can turn the roller by hand? AND before ya do any work --- DISCONNECT electrical and lock it out !!!!!
BillE.Dee wrote:Like CJ stated, I'd bet that is some type of alloy. Can you take that drive chain loose so you can turn the roller by hand? AND before ya do any work --- DISCONNECT electrical and lock it out !!!!!
Yeah they’re working on dis assembling it currently, then then they’ll bring just the bottom roller to my shop and I’ll see what I can do. I told them it’s too tight I couldn’t do a proper repair with the roller in place. It’s not that big of a job, take the top guard off with the motor and gearbox and then rollers are fairly easy to take out. I’ve taken them out before the replace cracked pillow blocks.
I’ll get a hold of the manufacturer first thing tomorrow then and see if I can find out what exactly the alloy is.
I did notice there seems to be an inner ring around the roller. That’s got me wondering. But it was kinda hard to get a good look at it behind those sprockets so I’ll have a lot better look once I get the roller out.
I’m guessing a lot of this is just from fatigue. The truss company specializes in big ag and commercial trusses, 60’ to 125 foot long clear span trusses. Being as they’re out in the country they can build and ship trusses that size where other residential roof truss builders in the city could never. But because of those long trusses they use a lot of much heavier gauge connector plates and hard lumber, a lot of 2400 grade yellow pine or Douglas fir which is 4 or 5 grades above your average framing lumber. You can barely hammer the connector plates into that type of lumber so they use hydraulic presses and then shove it through this finishing roller to finish. That’s an crap load of force on those rollers I’m guessing that’s 90% of the problem here.
cj737 wrote:Dodgy. I think you really need to ascertain the composition of the roller core and the outer face. It could well be the face is an alloyed steel that is not friendly to welding. The crack could be the result of improper pre- and post- weld treatment.
One thing is for sure, once I knew the material and treatment, I'd grind a V groove to remove the old repair and get into fresh clean metal. I would choose to lay a thinner root with even a 3/32 or 1/16 wire (stick or TIG) then fill from there. May not be necessary, but I would want to drive new wire in deep and fully, under the correct temps. Or it will likely just crack again, and again.
I don’t have a big enough tig machine, or 3/32 rods. I thought I’d run a hot 1/8” 6010 around into the crack after grinding everything out. Then pulse mig it from there
What I really like about pulse spray in this case especially is I can set my trim ways too high on the root pass, use a bit of a push angle and It’ll keyhole the crack ahead of me. Then stack 2 more passes with a reasonable trim setting. I’m pretty confident about the penetration even if it’s not my prettiest weld. I did struggle tho with it being round, and with such a hot weld the puddle is really fluid and wants to run down to the bottom. So I welded it up in short segments, turning the roller as I went. I guess the alternative would be to run more, cooler, smaller beads?
No I did just the outside drum. The didn’t want to take the roller press apart for me, so it was pretty much impossible to grind it all out in place. So I elected for adding more wedges, die grinding the cracks out somewhat and pad some 5/32 7018 over the cracks.
from what i'm seeing...as big and heavy as that drum is, I believe it is oil canning ever so slightly with all of the pressure they are putting on it. OR because of the shape of the truss, the pressure is starting on the center of the roll and radiating to the edge with a wobble.
BillE.Dee wrote:from what i'm seeing...as big and heavy as that drum is, I believe it is oil canning ever so slightly with all of the pressure they are putting on it. OR because of the shape of the truss, the pressure is starting on the center of the roll and radiating to the edge with a wobble.
99% of the forces are right over that weld, the trusses are lined up with the bottom edge of the roller and are fed straight out. Along the bottom edge of the truss is where the heaviest lumber and biggest, thickest connector plates are. The middle of the roll will see some force, maybe 1/4 as much as the right side, the left side sees virtually none. 16’ tall trusses (which is how wide the roller is) aren’t all that common and the peak connector plate isn’t very big usually. I guess because the top chord of a truss has more compression forces compared the the tension forces on the bottom chord which basically wants to rip apart. If it’s a free span truss that is. Hence the much beefier plates.
This picture is a bad example tho, it’s a fairly flimsy truss compared to what they normally make.
WOW !!!! jaywal, after you get that fixed and they run it again, can you have them put some wood under the roller where the trusses are angled to take the pressure throughout the roller ?? It looks like they are "tipping" the roller against the end of the tube and on your welds ... I'm also wondering IF the inner plates are broken loose
BillE.Dee wrote:WOW !!!! jaywal, after you get that fixed and they run it again, can you have them put some wood under the roller where the trusses are angled to take the pressure throughout the roller ?? It looks like they are "tipping" the roller against the end of the tube and on your welds ... I'm also wondering IF the inner plates are broken loose
That’d be my guess as well, the inner plates are loose. And since this is the driven side it sees a crazy amount of torque. You could hear grinding noises and klunking each time a large connector plate came through, and the roller flexed down enough that it wasn’t squeezing the plates into the lumber completely.
They did order a new roller when this one cracked a second time. So I’m just going to reinforce the new one once it does come. And scrap this one, maybe try to fix it again by someone with a big enough lathe and more skill than I have. $18,500 for the new roll tho. Sure would’ve liked to be able to fix it.
cut the end plates out ,, hire some skinny kid ,, teach him to weld and shove him inside...
have you been around when they operate that thing? There is definitely some bad flexing going on there.
BillE.Dee wrote:cut the end plates out ,, hire some skinny kid ,, teach him to weld and shove him inside...
have you been around when they operate that thing? There is definitely some bad flexing going on there.
Figure out later how to get him out eh?
I’d imagine, but they did have another older roller press that they’ll use till the new one arrives so they’re not using this anymore. Bugs me, because I hate not knowing why a repair didn’t last.
I’m pretty sure it’s from the torque on the shaft, the sprocket on the roll is 24” diameter, slightly smaller than the roll, and the drive sprocket on the gearbox is maybe 4” diameter, using double 100 chain. That’s a pretty significant gear reduction and it’s not a small drive motor.
can you work something out with them to investigate so they don't break the new one?? I can't tell by the pics if the original wedges and roller end are beaching from torque on the roller. They also might want to fit 2 trusses in opposite directions to take the pressure off the roller.
All the best.
Bill
BillE.Dee wrote:can you work something out with them to investigate so they don't break the new one?? I can't tell by the pics if the original wedges and roller end are beaching from torque on the roller. They also might want to fit 2 trusses in opposite directions to take the pressure off the roller.
All the best.
Bill
I do want to check it out, I’ll take the broken roll to my shop once the new one gets here. I have a really good relation with those guys, they’re less than a mile from our farm, I’ve done a lot of work for them. But I’m drawing blanks trying to think of something to make with a 5,000 lbs roller [emoji3]
I also want to reinforce the new roll before they even use it. Triple pass weld around the outside and add 4 wedges again. And I’m trying to convince them to add a VFD so they can slow the roller down for big trusses, so it eases the bigger plates in. It’s quite fast now, it pulls the trusses through at roughly a fast walk. Which they need for smaller (30’-50’) trusses as they can completely assemble 1 in under 2 mins. With a frequency control they could speed up again for the small stuff, but take it easy on the heavy stuff.
