[Erik Mannie]

I am making between 30 and 300 of these plates. These are used to bolt on the inside of a very fat tire for an electric bicycle in order to repel nails and screws. Each bicycle tire gets about 26 of these. This is about 9 added pounds so it is only for ebikes with a very large motor.

I am using 14 gauge (.080") hot rolled mild steel because I could drive nails through 16 gauge. I tried 5052 aluminum, but it was very soft. I am making prototypes now; the production plates will probably be titanium but possibly Kevlar.

Here is one plate. It has the curvature of a 5" diameter circle (does not need to be too precise). The main base is 3" long (tolerance here is a loose 1/8"). The plate has a protective overhang which is about 7/8" (1/8" too short or long starts to cause a problem).

The only precision tolerance is that they should each weigh about the same (for wheel balance). The closer they all are to the same weight, the better. The weights are standardized in the last step by grinding off some material.

[Erik Mannie]

So I am just learning how to weld and work with metal. I haven't even started welding school yet. My best idea was to start with a 4" X 4" sheet of the 14 gauge and round the edges on a bench grinder. Then I measure out 3" and cut off the end (that is for the overhang which gets welded up at an angle which also has a loose tolerance):

[Erik Mannie]

Then I shape the metal on a 4 1/2" pipe with a sledgehammer. I know it is low tech.

[Erik Mannie]

Now I have two curved pieces and I hand fit each one on a bench grinder. The overhang goes up at an angle.

[Erik Mannie]

Here is the overhang at the grinding wheel:

[Erik Mannie]

After I get an okay fit up, I remove the mill scale with a wire wheel and clean with acetone:

[Erik Mannie]

The best way that I have found to hold the two pieces together for the first tack is with painting tape:

[Erik Mannie]

For the rest of the following pictures, I had to create a gap because it was 3 AM and I didn't want to use my bench grinder and disturb the neighbors. Most of the time the fit up is better:

I clamp it down on the welding table and do one center tack on the inside, so that it will warp upwards. The upwards curve is beneficial for the product application because it keeps the plate away from the adjacent plate while the tire is flexing while in motion.

[Erik Mannie]

This is disgraceful. The filler there is ER70S-2. I really should have been working during normal hours and been able to take it to the grinder to make the adjustment:

[Erik Mannie]

So now it is tacked a bit, but I had to take the time to remove the painter's tape and clean off the adhesive with acetone:

[Erik Mannie]

More disgraceful filler rod used to fill the gap for the second tack. Like I said, usually I have a better fit up:

[Erik Mannie]

And the same thing for the third and final tack:

[Erik Mannie]

Here it is tacked with this "terrible fit up" example plate:

[Erik Mannie]

Normally, I run a TIG bead on the top. With this example, I had to do a root pass. I really should have posted a plate with a good fit up.

[Erik Mannie]

Remembering that I am new to TIG welding, I ran my best bead over the root pass:

[Erik Mannie]

If the bottom of the plate ("the other side of the seam") needs it, I will run a bead there, but the back side seemed to be good:

[Erik Mannie]

So clearly a skills deficit here, but I am signed up to go to Hobart in January. It really does look okay after grinding.

These go inside an ebike tire, so they don't have to look good, but the weight distribution should be as precise as possible.

It is kind of a strange situation because I am making the prototypes out of steel. Each design attempt needs about 30 plates made in order to be tested. If the final version were to be made of steel, I could make a fabrication process. Is it normal to just hack together prototypes like this? This design will probably evolve after the testing phases.

I'm not sure if it would have been worth it to have the flat pieces cut on a plasma table so that I could have enjoyed uniform weight and better fit up.

My main question is: would you guys and gals weld both sides of the seam? Isn't the theory that a good weld bead on top would suffice? To what extent does the additional heat of an extra bead (if deemed necessary) weaken the plate? The 14 gauge, as it comes from the supply yard, passes the destruction test. I wonder if my HAZ would survive a nail or screw being hammered through it. When I have a chance (I always work during the day), I will try to hammer a big, hard and sharp nail through the bead and HAZ.

[Erik Mannie]

I also wonder if I should have my SolidWorks brother-in-law make a drawing with the 5" diameter curve on the 4" X 3" sheet intersecting with a 7/8" X 4" sheet with the same curvature.

The 4 outside corners of the finished plates are always rounded so that they don't puncture the bicycle tube. I have been eyeballing the rounded corners on the bench grinder.

It seems that I should be making a drawing in SolidWorks or Fusion 360, having the pieces cut on a Plasma table, maybe bending those in a press (I have a 12 ton hydraulic press), and then just welding one bead on the top of the plate.

Final point is that the plate has a layer of yoga mat foam on the top, so it never gets close to the bicycle tube. The bottom of the plate, however, is right up against the thin bicycle tire so the bottom of the plate has to be pretty smooth.

These plates are bolted to the tire with two low profile M4 bolts and washers.

Thanks in advance for your help! This is a completely open project so if anybody wants to adopt my idea and improve on it, you have my permission. I have ebikes that need this extreme level of flat protection. The added weight of the plates is ridiculous, but being late to work is a serious problem for me.

[Erik Mannie]

I also wonder if I should have my SolidWorks brother-in-law make a drawing with the 5" diameter curve on the 4" X 3" sheet intersecting with a 7/8" X 4" sheet with the same curvature.

The 4 outside corners of the finished are always rounded so that they don't puncture the bicycle tube. I have been eyeballing the rounded corners on the bench grinder.

It seems that I should be making a drawing in SolidWorks or Fusion 360, having the pieces cut on a Plasma table, maybe bending those in a press (I have a 12 ton hydraulic press), and then just welding one bead on the top of the plate.

Final point is that the plate has a layer of yoga mat foam on the top, so it never gets close to the bicycle tube. The bottom of the plate, however, is right up against the thin bicycle tire so the bottom of the plate has to be pretty smooth.

These plates are bolted to the tire with two low profile M4 bolts and washers.

Thanks in advance for your help! This is a completely open project so if anybody wants to adopt my idea and improve on it, you have my permission. I have ebikes that need this extreme level of flat protection. The added weight of the plates is ridiculous, but flat tires make me late for work, and that is unacceptable.

[Erik Mannie]

I almost forgot my main question: I have settled on 85A max (no pulse involved although I tried it). I am using a #8 gas lens, 3/32" tungsten ground to 30 degrees with the tip ever so slightly blunted. I use 1/16" ER70S-2 and 20cfh straight Argon.

I have 1/16" tungsten on hand (all my tungsten is 2% lanthanated, guess where I learned that) and a good selection of #4 to #12 standard and gas lens cups and collet bodies. I also have plenty of 3/32" ER70S-2.

If I could get a good pulse setting, I could use the technique used to build bicycle frames. When I do that, it is 25cfh pure Argon, 1.8 PPS, 125A max, 25% on time, 10% background current for .049" 4130 Chromoly. I have done that a fair amount, but I could not find workable pulse settings for this .080" mild steel.

I am not experienced enough to know if I should even try a standard nozzle, what cup size to use or what gas flow I should use. I have been spending a lot of money on Argon for this project. I notice that I do a better job if I have a long stickout.

[cj737]

I’ll wade in with some impressions...

First, your gas setting is too high. You really only need about 12-15cfh for a #8 cup with a gas lens. 3/32 tungsten is fine, grind to a crisp point, don’t blunt it.

For the application, you do not need 2 passes of weld. Heck, probably 5 tacks is plenty strong if you are welding with an open root. Me, I’d consider welding the pieces together before bending or stamping them. Welding flat pieces together is much easier. You might need to perform a single, full pass though.

Your beads show insufficient heat, but likely due to using 3/32 filler. Drop to 1/16” using the same heat. Tack both ends without filler, then lay a single pass with the slightest weave across the gap. This should help you flatten the bead and burn it in better. I’d weld the “inside” of the part away from the rim so any irregularities in the weld do not disrupt the fitment against the rim. Leave that as bare metal.

Here’s another idea: use silicon bronze wire instead of ER70. It does cost more, but it welds faster because you use less heat, thus you create less distortion in your part. For the application, it will be more than strong enough too.

Last bit of advice- if you lay your first bead and it is humped up, uneven, and ugly, run a second pass without filler, just using the torch and “wash” the bead in. Use your pedal to avoid over-melting the weld, but enough to manipulate the weld. You’ll get better looking results and insure you’ve made proper fusion/penetration (as long as you don’t go nuts).

[Erik Mannie]

I’ll wade in with some impressions...

First, your gas setting is too high. You really only need about 12-15cfh for a #8 cup with a gas lens. 3/32 tungsten is fine, grind to a crisp point, don’t blunt it.

For the application, you do not need 2 passes of weld. Heck, probably 5 tacks is plenty strong if you are welding with an open root. Me, I’d consider welding the pieces together before bending or stamping them. Welding flat pieces together is much easier. You might need to perform a single, full pass though.

Your beads show insufficient heat, but likely due to using 3/32 filler. Drop to 1/16” using the same heat. Tack both ends without filler, then lay a single pass with the slightest weave across the gap. This should help you flatten the bead and burn it in better. I’d weld the “inside” of the part away from the rim so any irregularities in the weld do not disrupt the fitment against the rim. Leave that as bare metal.

Here’s another idea: use silicon bronze wire instead of ER70. It does cost more, but it welds faster because you use less heat, thus you create less distortion in your part. For the application, it will be more than strong enough too.

Last bit of advice- if you lay your first bead and it is humped up, uneven, and ugly, run a second pass without filler, just using the torch and “wash” the bead in. Use your pedal to avoid over-melting the weld, but enough to manipulate the weld. You’ll get better looking results and insure you’ve made proper fusion/penetration (as long as you don’t go nuts).

Very helpful advice! Thanks for saving me money in Argon expenses.

I have been using 1/16" filler. I tried 3/32" filler, but it was too thick. I am using 3/32" tungsten. Do people usually match the diameter of the electrode to the diameter of the filler rod?

I sure am glad that I only need to weld one side. I was concerned about the HAZ.

Do you have any recommended pulse settings for this application (.080" mild steel with 1/16" ER70S-2)? If I had to guess, I would guess 150A max with 1.8PPS, 30% on time and 20% background current. The vast majority of my TIG welding has been laywire pulse on .049" 4130. I was using a 1/16" electrode with a #12 cup and 25cfh for this.

I have been doing a lot of washing the weld beads. It would seem that I am afraid to use a lot of heat because I don't want to burn a hole in the base metal. Not that I mind fixing holes.

About 50% of the time I burn the edge even though I am being 100% conscious to taper down the current. I think this is because the piece has warmed up by the time I get to the edge.

I am just starting to work on this now for the next 5 hours. I will post pictures, hoping for improved results.

[LtBadd]

Do people usually match the diameter of the electrode to the diameter of the filler rod?

If I may jump in and give my opinion
No, the filler dia is matched to the weld you're trying to achieve.

Do you have any recommended pulse settings for this application (.080" mild steel with 1/16" ER70S-2)? If I had to guess, I would guess 150A max with 1.8PPS, 30% on time and 20% background current.

I don't, although I could give settings, IMO pulse "settings" can be a personal thing, you'll find good instructors that give a starting point.

For example if I knew I wanted to weld something around 75 amps, I could set the amperage at 125, % on time at 50% and background at 15%,(PPS at your discretion) I would then hold the torch on a scrape piece and initiate the arc and use full pedal while looking at the amp display to see if my max amps is about 75 amps, if not adjust one or more setting to your liking. The preceding is just an example, and the numbers I used arbitrary.

[LtBadd]

A second thought, if you want to learn what the different setting do, then change just one setting at a time and weld, but also hold the torch as described above and watch the amp display.

For example pick your settings and use a low PPS, watch the amp display, then move the PPS higher. (keep all other settings the same) Lets say start at <5 PPS then 15, then 30 then 50 or higher, or pick a similar pattern of your choosing.

[cj737]

I don’t really see a need for pulse with this application. It’s a useful tool for really thin stuff, but I would simply full pedal at a slightly lower amperage and match your travel speed to the weld needed. Don’t get hung up in too many things. Welding is an art, and each persons brush strokes produce their own art.

The longer the bead length, the less heat you’ll need at the end of the weld as the material has absorbed heat during the process. That’s where a pedal comes in handy.