3pt carryall



Note:  This design is intended for my personal use, and I make no warranty of fitness for a particular use.

The overall plan is to make a platform for carrying stuff in rough terrain, where I don’t care to run a truck.  I’d prefer to be able to carry it with the tractor. I’d like to carry a load in the 1000# range.  It’s designed for use with a medium sized 100hp, CatII hitch, 4000# lift capacity, tractor. Prior to drawing plans, I’ve decided on rough dimensions.  The platform will be approximately 7′ wide, and extend 50″ from the hitch.  (So, we’re talking something 84″long, 50″ wide).  Main structure will be 3″ square tubing.  A mix of 3/16″, and 11ga(approx. .125″), wall thicknesses. The first concern is whether the main carrying beams will support the weight.  A 50″ cantilever extending rearward from the hitch.  Three of them. To arrive at what I consider an adequate safety factor, let’s do a worse case scenario.  500# at the very end of each support beam. Plug in the properties of the beam.3pt carryall1Decide on the features of the beam.  How it’s supported, its length, and where the load will be concentrated. (click to enlarge)3pt carryall2Run a stress analysis. (click to enlarge)3pt carryall3 The beam itself is good to go.  The bending stress is well within 36ksi (A-36 mild steel specs).  The deflection is limited to approx. 1/3″. My concern is the moment at the connection.  Where the beam attaches to the crossbar.  Looking at both images, it’s apparent that the torque is 2000ft/lb in the stress analysis, and 25ksi in the beam support reaction analysis.  It’s a lot of force.  70xx filler material, to be used for the welds, should withstand the stress.  A small cover plate at the connection would be added insurance. The rotational force has to be countered.  So, it’s apparent that the crossbar has to be fairly strong to resist twisting.  My feeling is, 3/16″ wall tubing should be adequate.  This could be subject to change.  I sometimes look at something, and experience tells me to go heavier.  Of the 3 evenly spaced beams, the outermost 2 will have the greatest twisting effect on the tubing. Remember that we’re just analyzing ONE of the three beams which will share the load, and we’ve placed a maximum load at the extreme edge of the beam.  The “far edge” loading gives me a safety factor I can live with.  It’s very unlikely that you’ll have 1500# dangling on the combined edges of the three beams.  It’s more likely to be distributed along their entire length to some degree.

Preliminary Drawings


Some dimensions are not given.  This is because they’re subject to change, depending on the width you choose for your own use.

Nothing fancy, just something to work with, and allow me to cut the steel without having to stop and re-think measurements.  Drawings are a real time saver.3pt carryall5plans 3pt carryall 1These are subject to change as the job progresses, but give a general template.



All components are cut, and ready for minor machining, and assembly.

3pt carryall63pt carryall73pt carryall8The chop saw, when properly set up, will do a reasonable job.  It shines on tubing, but is a little off when cutting flat stock (The flat stock is held in the vertical position up against the fence when cut).  It can be off as much as the thickness of a pencil line.  Accurate enough for general fabrication, but not quite up to absolute precision work.

Even when being careful,  a chop saw wheel will deviate to some degree as it cuts through the metal.  It’s a matter of inherent flexibility.   This tendency can be reduced by using a steady down pressure, and NOT forcing the cutting wheel into the work.

It’s my feeling that chasing accuracy beyond this point is an exercise in futility, and simply throwing money at the problem.  You have to make some compromises.  For greater accuracy I tend towards using a plasma cutter.  But, it’s an expensive alternative.  Only justifiable if you have plenty of other uses for a plasma system.

To take into account chop saw inaccuracy, the pieces are bookended.  This matches surfaces to make accurate measurements  possible. 3pt carryall9Because they’re bookended, the pieces must match during the entire machining/welding process.  Numbers are stamped on the metal, and some measurements are scribed.3pt carryall10It’s necessary to permanently mark the metal so that some measurements are not washed off by the coolant.3pt carryall11

Sub Assemblies


For ease of handling, and welding, the sub assemblies were put together before assembling the main frame.

One of the problems with the small parts was the complex angles.  This leads to arc blow when stick welding.

Very often, it’s a real PITA when you’re welding something that has a lot of corners, or enclosed areas.  The magnetic field generated while welding with DC current leads to arc blow.  The arc will wander, blow BB’s, and generally ruin your day.  The only recourse, at times, is to resort to AC current.  AC is non-directional, and eliminates those pesky magnetic fields.  BUT AC IS GENERALLY NOT A GOOD ROD TO WELD WITH.  It can suffer from porosity issues.  Never used to, but nowdays seems to.  (Maybe it’s just me, but I’m thinking they’ve changed the formulation of the flux, or something.  Used to run like a dream up to about 3yrs ago).

For my money…AC can make for some real ugly welds.3pt carryall35Tis what it is, and you have to live with it.  The only way to make it run halfway decent is to shove that rod in there until it’s riding on the flux coating while welding.

If you’re in an area with no surrounding angles, it’s possible to run DC without arc blow if you weld towards the middle of the piece.  DC doesn’t like it when you weld towards “air”.  Ends of pieces, at the end of a bead, will suffer from arc blow.  Weld towards the middle from both ends.3pt carryall32The gussets on this assembly were beveled for a full penetration weld, and the corners nipped to allow for clearing the existing fillet.  In addition….you do not want to fully weld the inside corner of a gusset, it makes for a concentrated stress area.3pt carryall373pt carryall36 The ear assembly for the top link posed a severe distortion issue.  Welding something like this always does.  It will pull with the heat.3pt carryall33The piece is clamped in such a way to put pressure on it in the direction you want it to move.3pt carryall34Then the outer crown of the bow is heated, sprayed with water, and shrunk back to shape.  If you look at the picture closely, you’ll notice that the piece shrunk further than anticipated, giving it a reverse bow.  The piece was flipped over, and the same method used to recover the desired shape.  Then it was attached to the frame member.3pt carryall29The lift pin brackets were also attached to the frame members.3pt carryall30 As usual, once the pieces were attached to the tubing, there was distortion caused by the heat.  The tubing was then straightened.

Using a straightedge, the exact point of bend is isolated, and marked for heat application.3pt carryall31Heat is applied at the uppermost side of the bow, and to the web directly next to the bow.  NOT ON THE OPPOSITE FLANGE.  The heat is applied (a dull red), and then the piece is sprayed with water to rapidly cool it.3pt carryall26The weight of the tubing is insufficient to apply enough force, so some additional weight is added to the top of the bow.  (Note the crap piled on it).  And when everything cools, you get this.3pt carryall27Ready to go.

The tubing where the lift pins are attached is too short for application of weight, so it’s peened after heating.  Peening relieves stress, and allows the metal to shrink back to spec.3pt carryall28

Main Frame

I use a rudimentary jig to assemble everything.  Nothing fancy, just a simple set of trestles.jig1The pieces, the project sit on, are aligned by sight to be planar/flat.  It eliminates the need for a flat floor.  No leveling, etc.  It just has to be planar, and square.  The trestles are shimmed to bring them into spec.jig2This isn’t magic, it’s just plain common sense setup.  Your eyes are as accurate as most scientific instruments.  Flat and square, and you’re good to go. The frame is tacked, and welded, as it sits.  I think there’s no problem if someone has to flip it to weld it, but it’s easier to do it in place. All of the inside/outside corners are welded first while the frame sits on the trestle. 7018 is a beautiful rod for out-of-position welding.  It has a very nice smooth flowing puddle. For uphill welds…….Keep the rod at the face of the puddle, don’t go above it, and drag the puddle along as you progress upwards.3pt carryall183pt carryall23The most difficult welds were joining 3/16 to 11ga tubing (the end pieces are 11ga, while the pieces in the center of the frame are 3/16).  The heat has to be adequate to give good penetration on the 3/16 material, but it poses a problem with the 11ga (tendency to burn through).  Keep the heat focused on the thicker material, and let the puddle wash into the thinner material.  In this case, I had to run a keyhole.  The thinner metal is partially destroyed, and the filler material fills the gap as gravity pulls it down.  A bit scary, but you do need the proper heat to match the thickest material.  Always set your machine for the thickest material you’re welding.3pt carryall20 At this point, the main frame is almost complete.  After the butt welds are completed, it’s ready to marry to the floor frame.3pt carryall16


Tubing presents a problem when butt welding.  You’re presented with a large gap that has to be filled.3pt carryall39While this might not be a problem with MIG, or FCAW, it does present an issue with stick welding.  7018 is runny stuff, it’s not a fast freeze rod.  So I like to do butt welds with two passes.

First pass to close up the gap a bit.  The heat is focused on the shoulder of the tubing opposite the knife edge.3pt carryall40This weld was done with 1/8 7018 running at 115amps

The second pass closes up the gap.  Care has to be taken to keep from burning back the knife edge.  Do this, again, by focusing the heat on the shoulder of the tubing opposite the knife edge, and letting the puddle wash into the edge.3pt carryall41This weld was run at 115amps.

I felt that I was running a bit hot at this point.  It was requiring excessive manipulation to build the bead profile.  So I turned it down a bit.

First pass.3pt carryall42I’m running at 105amps here.  Notice the finer ripples in the bead.  This is due to being able to drag the rod without any manipulation.  It’s running cool enough to build a profile without excessive heat trying to burn through the tubing.  The amount of filler required to fill such a large gap requires that you run very slow to get enough melt-off to fill the void.  Run that slow, and heat can work against you.  I’m able to run slower, and deposit a greater amount of metal.

The second pass.3pt carryall43This was run at 115amps.  But this time, because I was running cooler on the first pass and filling a larger amount of the gap, I was able to run a nice straight drag with no manipulation.

Part of the floor frame was fitted, and tacked, later in the day.3pt carryall50From this point on, I’ll be mixing welding processes.  Where the 11ga floor frame joins the thicker material of the main frame (3/16) it will be stick welding.  Where 11ga is attached to 11ga, it will be self shielded flux core welding (FCAWs).

Floor Frame and Rails


Floor frame is complete.3pt carryall55The floor will be 12ga sheet metal, so it needs support in the middle of the span.  1″ 11ga square tubing is adequate.  No need to overbuild.  (It’s one of my pet peeves 🙂 )

For the “Nattering Naybobs of Negativism” (quote attributed to one of our past Vice Presidents) I went a bit further than jumping up and down on the supports for my load test.biggrin3pt carryall603pt carryall623pt carryall61biggrinEach piece of 1″ tubing will support 250# at HALF it’s failure value.  So we’re good.

When fitting the uprights for the rails, it’s a good idea to use a fixture (in this case a piece of tubing), rather than rely on pencil marks.  This gives you accuracy, and something to clamp the things to.3pt carryall54When tacking, the first tack should be a hinge. Clamp the piece tight to the base with no gaps on the bottom.  Using a square, line the piece up plumb, observe which way it has to be moved to be made plumb, and tack on that side of the piece.  This will be your hinge.  The second tack will be made on the opposite side after you tap the piece back into plumb.  This is an excellent way to make up for any errors that were incorporated into the cut by your chop saw.

A small note on “facing”.  Fitting anything which will run underneath a piece of sheet metal, etc.  Always use something to span your gaps.  In this case, I used a piece of tubing.  Clamp it to the surface on which the sheet will be applied, then fit your crossmembers.  Now you’re guaranteed a flush, flat, surface later on when you put the floor on.  This applies to anything that will provide support (in this case both the 3″ tubing, and the 1″ tubing).  Seems sort of a stupid reminder, but it’s not so stupid when the sheet metal doesn’t fit properly down the road.3pt carryall53After any finish welding is complete at a given stage of the build……….check everything for square, and flatness.  Flatness/planar can be eyeballed.  Sight it from a corner.3pt carryall56Gee, how did he get all that stuff to line up square?????

DIAGONALS.  Always measure your diagonals.

We have a right triangle (the angle of the floor frame, and the main frame).  Don’t have four corners in this case.

To get your diagonal, measure the legs, find an online right triangle calculator such as, plug in your measurements, and you’ll wind up with the required diagonal measurement.  I wound up lucky, and was only 1/8″ out of square.  Some heat, and it’s all good.3pt carryall57When heat shrinking/straightening, always apply your heat close as possible to the end of the piece.  The slight bend will be less noticeable in this area.  And, the rest of the piece won’t have a slight bow in it.  Try to keep from overheating any welds.



To hold the sheet metal at the “back” of the frame, a piece of angle was added.  Because of its open design, as opposed to tubing, it gave more space for reaching the area to be welded underneath the floor.  Angle isn’t as strong as tubing, but the back edge of the floor doesn’t require the support the middle will.3pt carryall63The angle, as well as the tubing crossmembers, aren’t welded to the shoulders of the main/floor frame.  This is to keep the main frame members from pulling when the weld cools.3pt carryall71aWelds in the neutral axis of just about any structural shape won’t create a lot of distortion.  The shoulder of the tubing is actually part of the flange, and welds here will possibly cause bowing.

The floor is 12ga sheetmetal.  Slightly thinner than standard 11ga diamond plate.  Strong enough for the anticipated load.

The sheets were sheared by the supplier for width, then I cut them to the desired length.  The 2 pieces were trial fitted, and marked for plug welds.3pt carryall64Then the pieces were book ended, and notches cut with a grinder.3pt carryall65This insures proper matching of the holes to be plug welded.3pt carryall67This portion of the floor, where the sheets meet, is on top of 3″ tubing.  To insure that the edges of the sheet won’t lift up when the underside is welded, the plug welds are necessary.  Welds were every 4″ both on the plugs, and outside perimeter of the floor.  The plugs won’t be ground flush.  My feeling is there’s no need to.3pt carryall72The sheet was held in place with fixtures during welding.  This is where the attention to keeping everything flush/flat pays off.3pt carryall66Welding is done from the center seam out to the edges.  Plug welds first, perimeter welds second, and underside welds last.  Working from the center allows the sheet to flatten out on its own as the heat progresses towards the outer edge.

The underside welds were spaced at about 10″.3pt carryall74On something like an equipment trailer, I would have spaced the underside welds approximately the same as the welds on top of the deck.  For something like this, which will be relatively lightly loaded, 10″ is adequate.

Always sight any surface to check your work.  See if it turned out flat, or needs some tweaking with heat.  This turned out perfectly flat, as hoped for.  Do everything in the sequence outlined above, and yours should turn out the same.3pt carryall68Managing heat input is the key.

Now it’s ready for preliminary fitting of the side rails.3pt carryall70

Side Rails


The top piece of angle is clamped to the uprights, along with the expanded metal (3/4 smooth expanded metal).  A fixture is also clamped along the bottom edge of the expanded metal.The angle is just tacked on the outside face of the uprights.  AT THIS POINT DO NOT FULLY WELD THE TOP OF THE UPRIGHTS TO THE ANGLE.3pt carryall77The panel fits flush on the inside of the uprights because they have only been welded on 3 sides.3pt carryall753pt carryall76I made a design error here.  I should have machined some weep holes in the unwelded sides to drain any water that might collect in the tubing.  Hopefully, any collected water will drain out of the crack.

While in the above clamped position, the bottom of the expanded metal panel is then welded to the deck.  This is done from the middle outwards to the edges, both ways.  This allows the panel to pull downward evenly along it’s length.  Always keep distortion/heat in mind when welding.  The tubing clamped along the bottom of the panel provides a straightedge for welding the lower part of the panel.

After the lower part of the panel is welded, a strongback is clamped to the top of the angle to prevent distortion.  Now the uprights are fully welded to the channel.3pt carryall78The upper side of the expanded panel is still partially floating at this time.  It has visibly moved downwards due to the shrinkage of the welds on the lower portion as they cooled.  This is why the welding sequence is so important.

The strongback is moved to the backside of the angle, and the upper side of the panel is attached.  Again….start in the middle, and work out to the edges.3pt carryall79The final welds will be on the vertical edges of the panel.

Following this sequence, these results are possible.3pt carryall1013pt carryall1003pt carryall883pt carryall102All edges perfectly match the frame, and the ends of the panel fall where they should.  This is the cumulative result of all the prior steps taken to keep everything square, and flat, from the very first piece of metal cut, or weld made.  It all starts to pay off as the project nears completion.

The backside is cut from 12ga sheetmetal, then notched to fit over the frame members.  Both the floor, and back panel, are set up to clear any welds on the main/floor frame.  These are critical welds, and shouldn’t be ground down in any manner to fit sheet metal.3pt carryall84A nice clean line on the inside corner where back meets floor.Then a length of angle is notched to fit around the main frame, and then clamped in place for finish welding.3pt carryall1043pt carryall863pt carryall89Again, we have a nice clean line.  And the angle will reinforce the top edge of the sheet metal.

All that remains to be completed is the tailgate.3pt carryall103

The Tailgate


The tailgate kicked my butt! 🙂

1″ square tubing is about the worst choice for a frame of any kind.  Distortion is a total TOTAL nightmare.

Fitting, tacking, and welding was done on the usual trestle setup.3pt carryall106First set of welds for the basic frame bent the assembly like a banana.3pt carryall107The thing had to be straightened to take out bowing, and warping on all axis.  This was the most difficult straightening in the whole process.  I believe I had to heat shrink about 10-12 different places.3pt carryall1083pt carryall109The hinges were welded to the frame next.  I’m sequencing assembly to make it possible to deal with distortion at every step.  If the hinges were welded at a later point, with more things already welded to the frame, it would be near impossible to straighten.  And…you guessed it…more distortion.3pt carryall110And more straightening.3pt carryall111The hinges were interesting.  I usually don’t do much welding on thin material, and have trouble following the lines when doing a thin lap weld.  The hinges are probably around 11-12ga material.3pt carryall1053pt carryall104aOnly way I’m able to do this, and I don’t do it well, is to make a bead of a given width that overlaps the lap edge.

Because the material is so thin, I felt it had to be welded completely all-way-around.  The knife edge along the shoulder of the square tubing was welded, then flap wheeled smooth to make the corner.

The expanded metal was welded to the frame last.3pt carryall112Welding was started at the middle of the frame, and progressed out to both ends from that point.  Also, different from the side rails, the welds were made on opposite sides as they progressed toward the ends.  And another round of distortion, but manageable because it’s in only one direction.3pt carryall1133pt carryall114Finally, the top piece of angle iron was attached, and again straightened after welding.3pt carryall115

The fixturing during the fitup, and progression of the welding progress, was mind boggling to say the least 🙂  Lottsa clamps 🙂3pt carryall1273pt carryall128
Finally, it was time to weld the stupid thing onto the carryall. AND THIS HAPPENS.erniefp3pt carryall129erniefpGet out another box of wire, and put it in.  But, the new roll is Lincoln NR211-MP.  It runs different than what I’ve been using.  So now everything has to be adjusted again.  All the parameters I’ve been using are wrong now.  This stuff runs hotter.  But the puddle doesn’t flow out like the Blue Demon wire.  About 15 minutes of yelling, and jumping up and down.  The dog went and hid under the shipping container.  (For those who don’t know me……….I’m the anti wire guy.  I HATE WIRE WELDING, only thing I like is stick welding…….just grab a rod, throw it in the stinger, set your amps somewhere in the ballpark, and weld)  But I digress………………………Finally got it glued to the carryall.3pt carryall120
The latching system was sort of an on-the-fly thing.  I hadn’t really thought of how to make the latches.  I wanted something that didn’t require pins, which are easy to lose.

I couldn’t sleep, and laid there looking up at the ceiling.  No way around a pin.  Ooooomph!!

If I gotta use a pin, I’m gonna make doggone sure that it can’t be lost.

No dinky little jewelers chain.  Decided to use something more substantial.3pt carryall119Could probably anchor a boat with this stuff 🙂3pt carryall124A portion of the end of the link was sliced out to accept the pin, then the pin was inserted in the gap, and welded.  Completely moron proof 🙂  I might forget to zip my pants, but I’ll never lose that pin!!!

The Lines

If you made it this far, without being bored to tears 🙂 , I had a few closing thoughts.

Keep it square.  Keep it flat.  And do it all like it’s something more important than what you’re actually making (This wasn’t the Space Shuttle).  Developing good practices will pay off on the really important things down the road.

Step back, and sight everything as you progress.  It should all line up, and the surfaces match.  If you do, then the final sighting will be something you like.3pt carryall1263pt carryall1253pt carryall123
I’ve always tried my best to make welds that don’t give away the position that they were welded in.  My thinking is that you should be able to look at any joint, and all the welds should look like they were made in the flat position.  There’s a lot of people out there that are better welders than me, but I give it my best shot.3pt carryall80 3pt carryall81 3pt carryall82 3pt carryall83
Anyway, thanks for sharing the ride, and I hope it helps someone.



Any day above ground is a good day

Return to past projects

2 thoughts on “3pt carryall”

  1. Where did you find the “Moment of Inertia” and the “Beam Properties”calculators? I have seen them in the past, but cannot find them again.

    Thank You.

Leave a Reply

Your email address will not be published.

error: Content is protected !!