I'm not punching cleanly...

[Beanwood]

OP needs to follow the advice of a few posters that said "punch one hole at a time, use a template to locate them".

You cheeky b@&£er - Being the OP I have read all of the advice- most of it I’ve taken on board. Some I haven’t - Because I know what I want(or have) to do- and it’s not what has been suggested. II want to finish what I started- which I believe is possible- even if not to the standard you may be happy with.

(I would love to be at a standard you ARE happy with - but at 60+ and a non practicing engineering background I suspect I don’t have enough years left.)

With the help I get from here I might get close enough to meet my own standards

 

[bricol]

There's a reason why some of us keep saying you won't get decent results with multiple punch set up - you simply cannot get the precision needed - my mention of your £1000 quote is simply that to get what you want of multiple clean hole punches in one hit, that's what sort of spend is needed.

To go back to the only machine I had designed to punch one hole at a time through one component - although there were 24 stations on a rotating turret, components being fed in, the whole lot rotating, the punch in each station being driven down and through the component as the turret rotated, banging out 300 components a minute. I needed to drive the punch up to 200 mm down into the tubular component before punching through the base, so I used high preload linear guides. As far as I'm aware, it's still working in the USA. I know very little about the actual punching - but learnt a lot about the precision needed for clean operation and decent tool life. Setting the punch to die clearance, getting it even all round, was absolutely critical to a good result. done mostly by feel - some of the operators got very very good at it. I/we tried various ways to make that setting up less skilled with dollies and jigs and failed.

It was a rapid design, build, test, modify, test again and deliver - it was air-freighted to the states at some ridiculous cost . . .

Why?

Another company had the original order and tried to punch 12 at a time . . . spent a million dollars . . . failed. I assume they would manage the first 12 pretty well, less well the next 12, even worse the next etc . . . the penalty clause they had agreed to would have bankrupted the company owners if the customer had insisted on it. Just could not get repeated precision.

So a little different to your application, but similar precision needed. You simply aren't going to achieve it with drill bits, a rule and centre punch . . . and I'm saying that as the guy in our office who will use a bigger hammer, compared to my colleague who will use ever decreasing tolerances that no-one has heard of on drawings . . .

I don't know how you are maintaining alignment between the punch mount and die plate - has to be absolutely spot on for just one punch, and resist the forces trying to deflect the punch as it works, especially if the punch tip is at all angled.

I'm all for banging your head (not mine!) against a brick wall, but your poor wall is going to suffer . . .

 

[Screwdriver]

There's a reason why some of us keep saying you won't get decent results with multiple punch set up - you simply cannot get the precision needed -

I am inclined to disagree. Beanwood is punching a few holes in a tin lid ffs, not a forward aperture detector plate for the Large Hadron Collider.

My seat of the pants guesstimate tells me each punch would need to penetrate the surface by 2mm for a clean cut. There are 18 holes. Set out in three groups of six, each "six pack" would only need to be 1-2mm more advanced in the die holder than the next. A flypress would make six holes in tin even if the die plate was a flat lump of wood (plastic, rubber, urethane or hard lead would also be candidates).

So if each set is offset by 2mm and there are three sets, the maximum stickout for the first set would only need to be 8mm (so that the final set also has 2mm of penetration). Whatever the numbers are, having the punches offset means they're not all trying to make a cut at the same time.

I merely give this as an example because there is more than one way to skin a cat.

Another aphorism springs to mind: if at first you don't succeed, give up. Is that what we're saying here??

 

[Agroshield]

Whatever the numbers are, having the punches offset means they're not all trying to make a cut at the same time.

But within any group of six, there is no way that all six will touch the material at exactly the same time. Whichever punch touches down first will try to twist the die. You will never eliminate that no matter how much you beef up its stiffness to resist that twisting.

You can see the same problem with a paper hole punch when using it at the top end of its capacity. One of the two punches always penetrates the material first and makes the thing clonk and jump a bit due to the frame twisting.

 

[Screwdriver]

But within any group of six, there is no way that all six will touch the material at exactly the same time. Whichever punch touches down first will try to twist the die. You will never eliminate that no matter how much you beef up its stiffness to resist that twisting.

You can see the same problem with a paper hole punch when using it at the top end of its capacity. One of the two punches always penetrates the material first and makes the thing clonk and jump a bit due to the frame twisting.

Hmm, that's a good point. Actually hadn't considered that. Looks to me like the setup Beanwood has constructed is way over engineered for punching tin (!), plus I assume it will have at least a 3/4" possibly 1" mounting pin.

But you're right, I do assume a sturdy construction and a sharp punch...

 

[Beanwood]

There's a reason why some of us keep saying you won't get decent results with multiple punch set up - you simply cannot get the precision needed - my mention of your £1000 quote is simply that to get what you want of multiple clean hole punches in one hit, that's what sort of spend is needed.

To go back to the only machine I had designed to punch one hole at a time through one component - although there were 24 stations on a rotating turret, components being fed in, the whole lot rotating, the punch in each station being driven down and through the component as the turret rotated, banging out 300 components a minute. I needed to drive the punch up to 200 mm down into the tubular component before punching through the base, so I used high preload linear guides. As far as I'm aware, it's still working in the USA. I know very little about the actual punching - but learnt a lot about the precision needed for clean operation and decent tool life. Setting the punch to die clearance, getting it even all round, was absolutely critical to a good result. done mostly by feel - some of the operators got very very good at it. I/we tried various ways to make that setting up less skilled with dollies and jigs and failed.

It was a rapid design, build, test, modify, test again and deliver - it was air-freighted to the states at some ridiculous cost . . .

Why?

Another company had the original order and tried to punch 12 at a time . . . spent a million dollars . . . failed. I assume they would manage the first 12 pretty well, less well the next 12, even worse the next etc . . . the penalty clause they had agreed to would have bankrupted the company owners if the customer had insisted on it. Just could not get repeated precision.

So a little different to your application, but similar precision needed. You simply aren't going to achieve it with drill bits, a rule and centre punch . . . and I'm saying that as the guy in our office who will use a bigger hammer, compared to my colleague who will use ever decreasing tolerances that no-one has heard of on drawings . . .

I don't know how you are maintaining alignment between the punch mount and die plate - has to be absolutely spot on for just one punch, and resist the forces trying to deflect the punch as it works, especially if the punch tip is at all angled.

I'm all for banging your head (not mine!) against a brick wall, but your poor wall is going to suffer . . .

Happy to agree - I will probably never get the results that you'd consider acceptable. And if I did - I'm sure it couldn't be repeated again at 30/min.

But I don't WANT to manually do 20 holes (My nirvana is actually 12 around the outside 8 on the inner circle) manually - I've tried previously. I want to spend all the time upfront getting a repeatable pattern sorted if I can. If I get an order for 40 - I'd like to be able to produce them in minutes - not tens of hours. However, I certainly don't have £1000000 to spend on it to get a total return over 5 years of possibly £1,000. But yes, they also have to be neat and tidy - and up to a standard I could stand by - which these aren't)

It was your posts that made me consider the alignment issue - and I have to find a better way of doing it.
I'm fortunate that I am not beholden to anyone else - and my volumes in a year would be less than yours in 2 minutes.

Hmm, that's a good point. Actually hadn't considered that. Looks to me like the setup Beanwood has constructed is way over engineered for punching tin (!), plus I assume it will have at least a 3/4" possibly 1" mounting pin.

Thanks @Screwdriver Yes, this is probably vastly over-engineered (in my opinion) for punching some tin. It's a 1" mounting pin.
No workshop time since Saturday morning - so no chanc eyet to rework anything else yet.
Roll on next weekend.....

 

[Erie Fred]

Yabutt !

 

[Screwdriver]

@Beanwood looking back at the pictures, I don't think you're a million miles off getting the result you're after with the setup you have.

Sharpening up and shortening the punches has been discussed to death...

The face of the die plate just needs a good linishing, 30 minutes on a flat plate with emery should sort that.

As for registration, I see how you are loading the workpiece which makes fitting guide rails a chore. Instead you could have two or three extra long punches which (carefully) enter the die first and register the rest of the (shorter) punches before you slam the handle. A decent flypress should be able to maintain the registration tolerance you need.

 

[ukracer]

@Beanwood looking back at the pictures, I don't think you're a million miles off getting the result you're after with the setup you have.

Sharpening up and shortening the punches has been discussed to death...

The face of the die plate just needs a good linishing, 30 minutes on a flat plate with emery should sort that.

As for registration, I see how you are loading the workpiece which makes fitting guide rails a chore. Instead you could have two or three extra long punches which (carefully) enter the die first and register the rest of the (shorter) punches before you slam the handle. A decent flypress should be able to maintain the registration tolerance you need.

It's frustrating as I would love to try and make one here. If nothing else to prove of disprove my vision would work.

I visualise two round circles of steel.
Holes drilled in using a rotary table through both. Them reamed
Smaller holes in part of the one circle to allow for a plug weld on each end of the pins and accurate length of pins.

A further plate with 3 locating pins to line up the bottom of the die with the 1 inch bar of the top punch fixed to register top to bottom.

 

[Beanwood]

@Beanwood looking back at the pictures, I don't think you're a million miles off getting the result you're after with the setup you have.

Sharpening up and shortening the punches has been discussed to death...

The face of the die plate just needs a good linishing, 30 minutes on a flat plate with emery should sort that.

As for registration, I see how you are loading the workpiece which makes fitting guide rails a chore. Instead you could have two or three extra long punches which (carefully) enter the die first and register the rest of the (shorter) punches before you slam the handle. A decent flypress should be able to maintain the registration tolerance you need.

Thanks - I'm going to remake the die plate. The holes are just a bit too sloppy. I'm planning to ream with 4mm and 5mm, I just need to get the drill sizes correct first.

You're right about the way I've constructed it making it difficult to have guide pins - but once set, it does indeed 'stay put'.
I have (On my smaller version) left one pin sligtly long to allow for initial registration when placing them into the press. On the current version, they are are the same length - which I may change. It's also quite tricky in current form to get reasonable springs on the stripper plate....

Smaller holes in part of the one circle to allow for a plug weld on each end of the pins and accurate length of pins.

I have tried the 'welded plug' on an ealier version. It's difficult in my hands to penetrate enough of the silver steel and the punch plate to make a solid and reliable connection - and the heat from the weld causes the (In my case silver steel) to go brittle. Maybe I could solve that someway. The other disadvantage is getting a damaged pin means grinding out and re-welding. That re-welding caused additional heat, and my 6mm mild steel punch distorted slightly...

I now hold the pins in a blind hole (25mm thick punch plate) - held in place with a 5mm grub screw . That works very well, and I can change to improve pins very easily to polish or replace damaged ones.

 

[Screwdriver]

You're welcome. With reduced penetration, you will find that the puches will come out a LOT easier. With such thin material, the longer punch penetration can lift the material far enough to rack in the hole. You may find even a thin sheet of rubbery material or a sturdy foam would release the part. Small 'O" ring would do it with sufficiently shallow penetration.

Agree that the flypress itself is as good a registration system as required once set up correctly.

 

[ukracer]

Thanks - I'm going to remake the die plate. The holes are just a bit too sloppy. I'm planning to ream with 4mm and 5mm, I just need to get the drill sizes correct first.

You're right about the way I've constructed it making it difficult to have guide pins - but once set, it does indeed 'stay put'.
I have (On my smaller version) left one pin sligtly long to allow for initial registration when placing them into the press. On the current version, they are are the same length - which I may change. It's also quite tricky in current form to get reasonable springs on the stripper plate....


I have tried the 'welded plug' on an ealier version. It's difficult in my hands to penetrate enough of the silver steel and the punch plate to make a solid and reliable connection - and the heat from the weld causes the (In my case silver steel) to go brittle. Maybe I could solve that someway. The other disadvantage is getting a damaged pin means grinding out and re-welding. That re-welding caused additional heat, and my 6mm mild steel punch distorted slightly...

I now hold the pins in a blind hole (25mm thick punch plate) - held in place with a 5mm grub screw . That works very well, and I can change to improve pins very easily to polish or replace damaged ones.

I had thought of that way. In fact I used bolts to hold parts in a jig I made.
They wanted it quick so ended up welding it. Worked fine both ways but easy to change parts when bolted.

 

[Beanwood]

You're welcome. With reduced penetration, you will find that the puches will come out a LOT easier. With such thin material, the longer punch penetration can lift the material far enough to rack in the hole. You may find even a thin sheet of rubbery material or a sturdy foam would release the part. Small 'O" ring would do it with sufficiently shallow penetration.

Agree that the flypress itself is as good a registration system as required once set up correctly.

Now you have me thinking! I have some small O rings. I am indeed experiencing the ruck you mentioned- I could experiment with adjusting the depth of press action to reduce the penetration to 3mm. Quick and easy to test

 

[Theblackbox]

This is a bit of a funny site but lots of useful info for punching forming etc all the calculations should be on there.

Press Dies Tutorial | Technical Tutorial - MISUMI

MISUMI's Technial Tutorial site empowers engineers and designers in the Factory Automation, Press and Mold industries to improve the efficiency of their design through valuable, in-depth tutorials.

www.misumi-techcentral.com

I would definitely be contacting someone like Berger tools for pre-made punches and dies. Then all you need to make/get made is a holder for them both to go into a bolster (used are available on eBay etc for good price)

 

[Theblackbox]

just to add one thing I don’t think has been mentioned if you do want to make your own dies have you added a taper relief to the back for the slug to fall out this will also make a difference.

 

[Beanwood]

just to add one thing I don’t think has been mentioned if you do want to make your own dies have you added a taper relief to the back for the slug to fall out this will also make a difference.

Thanks- yes on the smaller one but not this one yet- I’ll add it to the list

 

[Theblackbox]

Thanks- yes on the smaller one but not this one yet- I’ll add it to the list

This is front and back of a 5.5mm die that fits my iron worker so you can see what I mean. For reference these are about £12 each! Think generic Berger tools dies are not as much as that!