Boosting bottom

microgen / iStock / Getty Images Plus

Question: I need some help understanding 90-degree tooling. Our guys on the floor will use a 0.500-in., 90-degree die with a 0.030-in.-radius punch on 0.108-in.-thick galvanized material. This allows them to smush the material. They bend the part past 90 degrees, and then once they hit it hard enough, they can bring it back to 90 degrees. In order for them to even get the tonnage, this requires they change the tooling from 30 tons per foot to 100 tons per foot.

So, I wanted to test the 8x rule—that is, using a die opening that's eight times the material thickness. I went out on the floor and used a 0.750-in., 90-degree die and a 0.060-in.-radius punch. I just brought the material to 90 degrees. I didn't try to squeeze the tooling past its capabilities. What I found is that even though I am using 90-degree tooling, the angle is still bouncing around roughly 0.7 of a degree. I was expecting since we are using 90-degree tooling that my 90-degree angle would hold. How can we make these 90-degree angles not vary so much?

Answer: Let's start with what your guys on the floor are doing. Your description sounds like your crew is coining rather than bottom bending. What's the difference? In simple terms, coining is where you descend the punch to a less-than-material-thickness position. For example, if the material is 0.108 in. thick and you force the punch nose to any position less than that, you are coining. The material will be thinned at the point of the bend, and the punch's profile will be clearly visible as you "smush" the material between the faces of the punch and die (see Figure 1).

You did not state the angle of the punch you were using. However, from the process description, I’m guessing you’re mating a 90-degree punch angle with a 90-degree die angle.

Bottom bending occurs about 20% above the material thickness (see Figure 2). Using the data from our example, the punch's lowest point would be around 0.129 in. above the zero point, as measured from the bottom of the die (see Figure 3). Bottom bending only works when forming lighter-gauge materials, 16 ga. and thinner. It matches the punch angle to the amount of springback present in the material, giving you angular clearance between the punch and die.

You stated that you had to increase the bending tonnage from 30 to 100 tons per foot. I am not surprised, as a 0.5-in. die opening is on the small side. Unless you need to use a die opening of that size to, say, catch a flange or stop a feature from pulling, you might want to consider a larger die opening.

A large amount of tonnage can permanently damage your press brake either by embedding your tooling into the bed and ram or causing ram upset, the permanent bending of the ram and bed caused by exceeding the press brake's centerline load limit. By the same token, the 0.75-in. die opening is a little large. You should be using a 0.625-in. die.

Your operators are using 0.032-in. punch radius, and you tried using a 0.062-in. punch. Here, the 0.032-in. punch nose radius is bending the material in a sharp relationship to the material thickness (see Figure 4). When you bend sharp, bend angle variations increase. The sharper the nose radius is to the material thickness, the greater the amount of variation you will encounter from part to part.

As a rule of thumb, bends turn sharp when you form an inside bend radius that's about 63% of the material thickness. Applying this to your example, your 0.108-in.-thick material turns sharp at an inside bend radius of about 0.068 in. When bottom bending or coining, the punch nose determines the inside bend radius, and your operators are using a punch nose that's 0.032 in. They’re creating sharp bends, which can lead to increased angular variation from part to part, especially for those who air form. It also means that your choice of a 0.062-in. nose radius is better for this project.

Narrowing the die opening (width) to form the same material thickness increases angular error, the total required tonnage to form, and friction between the material and the die shoulders (see Figure 5). And if you’re bending with a sharp inside radius-to-material thickness relationship, you’re increasing this angular error even further. Note that when you are truly coining, almost 100% of angular error will disappear to what you describe as "smushing." Coining hits the material with so much force that properties like grain direction and springback muddle and ruin the integrity of metal at the bend. Nonetheless, there will still be some variation in the angle still present.

FIGURE 4. When you bend sharp (an air bending application is shown here), the punch nose can crease the center of the inside radius. Note, however, that you can still bend sharp and not leave a crease behind. Never use a crease as your only indicator of a sharp bend.

One more thing you always need to do, especially when bottoming or coining, is to ensure that your punch is completely centered in the bottom die. If you find your tooling isn't centered, make sure that the die is not bowed or bent. An off-center tool setup can also cause or increase angular variations from bend to bend.

I assume that your press brake and tooling are in good shape and relatively up to date. If so, the machine and tooling should not be contributing to your problem.

Most variations encountered in bending relate to the material and are enhanced by poor tooling choice. What do I mean? The material is everyone's weak point, and here is why. No two pieces of material are the same, from part to part, sheet to sheet, and even batch to batch.

Just look at a few of the variables, all related to tolerance. A 16-ga. sheet has a thickness tolerance of 0.014 in. In other words, a material thickness of 16 ga. can be anywhere between 0.053 and 0.067 in.—and still be called 16 ga.

That's just the thickness tolerance. What about the strength properties? A36 steel has to be at least 36,000-PSI yield strength to be called A36, and its tensile-strength value can vary from 58,000 to more than 79,000 PSI—but again, it all qualifies as A36 material. In fact, every possible descriptor has a tolerance or an allowable amount of variation.

With all this said, back to your question: How can you minimize your angle variation for those 90-degree bends? Change your die opening, bottom bend with a 1/16-in. radius punch, rather than your 0.032-in. punch, and you should be just fine.

To be honest with you, if your total variation is only 0.7 of a degree, that is about the best you could hope for in a production run. Sure, you could apply some of the concepts I’ve discussed, and you could get the variations down to a half of a degree from bend to bend. For bend angles with less than 0.7 degrees of variation, you will be looking at a lot of hand work to make these parts perfect. Nonetheless, if it were me, and I had only three-quarters of a degree variation from part to part, and that was all I was seeing, I would be a happy person.