I got a good deal on unused 24"x32" 1/4" plates at scrap price, so they'll be used for a number of things on this project. Here two have been butt welded and are being trimmed to serve as the new gas tank skid. We've only welded one side so far -- single pass V with the little MIG box. The other side will wait until we get a chance with the big Miller.

Skidplate dimensions are 37.5" across by 32", basically outside of framerail to outside of framerail, and #4 crossmember back to lap 2" under the new bumper.

Even with only about 50% weld penetration (a single 1/8" MIG bead), the plate still has no trouble holding JB up. We'll be adding at least two more 7/16" Gr8 bolts about 1/4 of the way back from the #4 crossmember. Right now the plate is supported temporarily using the old hardware, which includes four 3/8" Gr5 replacement bolts from the front skidplate on the #4 crossmember, and six lousy quality 7/16 Gr5's at the rear of the framerails (formerly used to attach the old hitch receiver).

New bumper endplates cooling after being cut out of 1/2" plate with the torch. Lots of cleanup still required. These plates will cap off the new 4"x6" tube bumpers and slide inside the back of the framerails, with (per side) three 9/16" Gr8 bolts in the original bumper holes, and three 9/16" Gr8 bolts at the narrow end forward of the shackle.

I still have to check numbers on what these frame attachments will take, but when I checked the tear resistance on the three original bumper bolts alone, they were good for around 9,000lb. That's with no safety factor -- at 9k, they rip out of the frame. The problem is actually frame strength, not fastener strength, and the three bolts forward of the shackle should be good for a lot more than 9k. I'm hoping to be able to safely use a 1" or 1-1/8" clevis back here for 14k-17k per, without having to rely too much on the skidplate and top plate welds for reinforcement. We'll see how many bolts we'll need to add up top to make 17k work safely.

Putting a square peg through a round hole ...

JB drilling the end plates.

Gussets tacked in place, one test bead run.

Lots of spatter and not the best penetration, but the remainder of this piece is overengineered. I'm still learning to work with thicker sections with this welder -- it's small, so heavier sections require *very* good joint preparation and technique. This bumper is a less critical item -- it's designed for a 25,000lb pull even with mediocre welds factored in, and hence will be *far* stronger than the original hitch receiver it replaces either way. However, it gives me a chance to practice before doing the cab crossmember, which is much more critical.

Pretty good heat affected zones on the backside, but the tack in the foreground is ugly as hell.

Backside cooling down after finish welding.

Business end of the bumper. Welds around the receiver tube are three pass fillet, chain plate is filleted to bumper on both sides and to tube where they contact.

I'd been fighting my welder the whole time, and like an idiot couldn't figure out why I was having trouble with it. Notice the complete lack of spatter on the last two welds -- after fiddling with wire speed, gas flow, travel rate, and replacing the contact tip, I *finally* realized my wire spool was binding just enough once a revolution to slow down the wire speed, break the arc, and throw spatter all over. Problem solved. Finish welding the backside would've taken half the time if I'd figured that out beforehand instead of having to fight the tool.

Time to make the final cut to length so we can mount the bumper between the endplates. Daddy's got a new toy. Sure, we had to modify the bandsaw a few times before we could get it to work right, but at $180 for a new working band saw, I don't mind doing a little re-engineering to make things work the way I want 'em to.

Tacked in place and ready to be unbolted for final welding. Right now it's secured by half the endplate bolts and five tacks on either end. Took fifteen minutes worth of measuring, checking, adjusting, rechecking, etc. to get it where I was dead certain I wanted it and it was lined up. Once it's finish welded, it'll be time to make the top plate that covers the back 8" or so of frame, and make the small 1/2" doubler plates for the end plates that will reinforce the shackle attachment points.

Finish welds on the end plates.

All done except the top plate and the reinforcement around the shackle points. Oh, and still need four more holes in the frame and the brackets to support the back of the skidplate. Maybe not so done.

With a little cleanup, these two rough ovals will become the reinforcement for the rear shackle points.

A "little" cleanup. New toy makes short work of rough edges on a pair of plates that total 1" when clamped together.

Plates are drilled, then beveled at the ends and tacked in place ...

Followed by finish welds. Single pass 5/32" or so fillets are more than sufficient for the required strength ...

... which is good, since that's the maximum capacity of this welder in a single pass. More strength could be obtained by proper multiple passes if necessary.

The infill at the end. I did screw up doing the bevel on the opposite end, and had to grind out about 1" worth of weld ... but that's actually a good thing, because it gave me a chance to check my penetration. These ends are solid.

Two endplates, plus four tabs to help support the skidplate. Then we quit on the truck for the day so we could go pull down a tree with the F-250.

I was asked for calcs and an illustration explaining why I said the factory mounting holes in the frame were insufficient -- these are the drawings of my first try at a bumper mount, and a set of calcs illustrating why that mount would not have done the job.

Illustration of the block shear failure area, and necessary dimensions.

Illustration of the rotation failure dimensions.

Design for a simple removable winch mount, much stronger than receiver mount and without the clearance issues.

We had a discussion on T4W about doing big jobs with a little welder. Here's an illustrative series of a full penetration butt weld in 1/4" plate using a welder that gives me 100A at most. It's not as pretty as it could be, but more practice on large sections will help there.

Start by cleaning up the edges and doing a *good* job of beveling both plates back about 45*.

Get the pieces lined up straight and clamped in place. Remember the typical MIG guidance about not having a gap larger than the diameter of the wire in your fitup? Throw that out the window. The two little aluminum pieces you see are sheetmetal clamps that don't have enough range to clamp the 1/4" plate, but the clamp tongues make a good gap gauge at around 0.040" or so while I clamp the rest of the plate down to the two 1.25" square bars underneath with four clamps each.

Then strip back the paint at the weld area, and clean. That filth on the paper towel was from one swipe with alcohol, one of my favorite cleaners. Obviously worth taking the couple of minutes necessary to get it clean, you'll be thankful later.

Tack it up ...

Then connect the dots. In this shot, I've just finished the middle section, as you can see. I had to pause here to move my bars underneath just to make sure I wouldn't accidentally weld them to the plates.

If you want decent penetration, don't ask the welder to do more than it can. Going slow just piles up a big bead, it doesn't get you any deeper. This is about the maximum capacity of this 110V welder. Run long beads when you can, the more often you stop the more cold starts you'll be grinding the backside of and re-welding. With this gap and bead size, I was only getting full penetration all the way through the backside after about 2" of travel, once the plate got some localized heat in it. You obviously have to be prepared to compensate for this with your joint design and method.

Run the second bead on the right side. My garage was a high traffic area at this point, so even with screens up I was stopping severy 2-4" so someone could come through.

Run the lefthand bead. Fewer stops this time, too, which is good since it's easy to trap small pockets on the last pass if you don't carry decent heat through. As you can see, no gaps, but none of the big pipe of steel you'd get if you were travelling too slow. You should be seeing about three seconds to the inch or so here.

Now, flip it over and get a look at the backside (I've already given the plate time to cool and cleaned the paint off here). As you can see, still an average depth of 1/32" or so that got no penetration, almost 1/16" in places where I had a cold start. This was expected and easy to fix.

Tilt your 4-1/2" angle grinder up 45* and V out the backside. Start with the worst spot, and open it out until any sign of a gap has disappeared, and you have a clean 90* V.

A single fast pass will finish the backside. I probably could've skipped these last steps if I were leaving the 1/16" or so of weld bead built up on the front side, though the small gap on the backside would still have lead to a stress concentration that would've formed the first crack if I'd bent this plate to failure.

But I'm grinding this plate smooth on both sides, because it's highly visible.

Where's the weld?

I need a 2-1/2" flange at the backside of this plate, so we decided to see if we could clamp itt, heat it, and fold it over. I didn't know whether the #4 tip would be enough heat to do the job on this long a bend.

Not a chance! Of course, if I'd known last weekend how Dave was doing his bends I could've done that, but for some reason that had never occured to me. Good idea, though. In this case, once I got away from the corner, the plate soaked up too much heat to bend, so now I've just bent up the flange area. I'll have to cut it off and just weld on a flange that isn't warped. Lesson learned.

Alright, here's the forward side of our new reinforcing flange, a 2x2x1/8 angle. These PL2x2x3/16 plates welded to the end are reinforcement to both help prevent buckling, and to help me clamp the angle square against the bottom of the winch plate to weld.

First pass down the root of the fillet complete. Honestly, I could probably have left it like this.

I also discovered that even with a good condition, new 20A breaker, all of 2' of 12AWG from the breaker to the outlet, and the welder plugged straight into the outlet ... the breaker still likes to trip after about 20" worth of bead at full output. I think I must've tripped the breaker five times putting this flange on.

Three more passes complete the fillet. Trying to keep it hot and fast as I can.

Another two passes to finish the weld of the endplate to the winch plate. Note the little bit of porosity at the end of the long weld on the angle -- only place I ran into it, I backed off a little far on the gun as I reached the end of the weld.

The endplate is also welded to the angle on the inside, so it didn't require quite the amount of welding on the outside. I don't intend to fillet weld the angle to the plate on the inside, so I need as good a weld as I can get on the outside side I can get to.

Opposite endplate, likewise three passes below and one on the top.

The whole stiffened winch plate assembly, flipped upside down end for end so I can clean the bottom. I'll butt weld the top to the bumper, then remove the assembly, flip it upside down, and fillet underneath.

Back where it should be, with areas to be welded now cleaned of paint and grime. Just needs to be clamped in place.

After screwing around with the little MIG welder all morning, I decided I should really use the welder big enough for the job for these critical welds. My stick welding is pretty mediocre, but I think I'd rather risk having to grind out a few ugly spots and re-do them than spend all day trying to get one good weld out of the MIG box on something this heavy.

Still haven't chipped the slag from the last rod. Not the prettiest thing in the world, but I could've spent all day with the MIG gun trying to get something this solid.

I really need to spend a lot more time practicing with the big arc welder. I think I'd end up getting things done a lot faster, in the end.

I've done worse, I only had to grind out two nasty areas and re-weld them.

One of the areas I ended up having to grind porosity and run a second pass on. I ended up deciding to do two passes on most of the run, because I had fairly large gaps and got a little undercut. Big crater at the end of the section I rewelded, top right, but it's all proud of the surface, and I'm content it will do the job. I probably wouldn't even need to fillet the backside, but I'm going to anyway.

Another section of the area where I did two passes. Still need to work on my rod control and get a more even bead.