Need help with a load cell application 4

[ak369]

I am looking for a particular load cell to use for an application in which the load cell will be used to determine the force profile inside of a chute that it is sent through.

I will try to be as specific as I can.

In the company I work for, there are presses which make parts, that in turn come out of chutes where they are collected.
These parts often end up bending as they are made out of steel.
In an effort to determine what is causing these parts to bend, we decided to try and determine the force profile inside of the chutes.
Someone recommended that I use a load cell, in which the load cell will be sent through the chute filled with parts in order to determine the force profile on the inside of the chute. The parts where the force profile is highest is most likely where the bent laminations occur.

The problem is, I barely know anything about load cells, and I need a particular load cell for this application. Essentially, it should be able to fit through a chute (I can provide the diameter if requested but it should be able to fit a load cell), and also help determine the force profile within the chute..

If anyone has any experience using load cells, please let me know.

I was checking out A&D

http://www.aandd.jp/products/weighing/pdf/loadcells.pdf

and the LC-5223-T005 load cell seemed fine. The presses are typically 250 or 300 tonnage, but this one had a 50 kN maximum capacity which was the highest on their website.

I think the beam load cells looked the best but which type or particular models would you recommend using?

 

[rb1957]

I don't think a loadcell is what you need.

You say "Steel" pieces, I assume in Normalised condition for forming.

It seems "odd" to me (in my ignorance) that steel pieces can be deformed easily, so I think ...

1) what speed are they being fired into the chute ? can you redue this speed (maybe absorb some KE on some baffles?) ?

2) how stiff is the chute ? Is collision with the chute creating the damage ? Can you make the chute walls more pliable ?

3) is there something in the shape of the work piece that makes this deformation a likelihood ??

4) Is this new production, or has this part been in production for a while and this is a new problem ? if so, what changed ?

another day in paradise, or is paradise one day closer ?

 

[MikeHalloran]

Maybe you shouldn't be using a chute, given that your laminations must be rather delicate.

It might be better to set up the (guess) blanking presses to accept feed material as a strip,
AND LEAVE THE STRIP INTACT when exiting the tool, carrying the laminations in the strip by means
of fragile webs that connect the product and the strip. Then the laminations can be broken out of the strip when it is time to use them.

The technique is used for retaining parts within a big sheet that goes into an NC punch press, and makes a lot of individual parts from the sheet. Once you've completely punched out the first part, you don't want it rattling around to get jammed in the punch working on the other parts. So you leave at least three webs holding each part to the sheet that it came from.

The trick to separate the parts: the operator lifts a corner of the sheet and shakes it; all the parts fall onto the table, and the operator then lifts the now skeletonized sheet away, and picks up the parts.

For steel of 1/16" to 3/16" thickness, a web between part and sheet of about .003" works ok.
No, you don't try to punch out a straight web .003" wide x 1/4" long.
You typically outline the part with a 1/4" punch, overlapping the hits to keep the finish acceptable, except at the webs, where you make the punched holes miss each other's tangent by .003" or whatever
you choose.

In a dedicated punch, you'd do it in a different way, but design the punch and die so that parts are supported and retained by short, narrow bridges of material.




Mike Halloran
Pembroke Pines, FL, USA

 

[Jboggs]

If you still want to use some active measuring technology, I think what you really need is an accelerometer instead of a load cell. Load cells are usually built assuming they will be stationary, thus making connections to power supplies and readouts convenient. But this could be difficult with a device "falling" through a chute. Also, they are usually uni-directional. They are usually intended to measure the forces applied to a surface in one direction.

On the other hand, accelerometers are built for motion. Its implied in their very name. Some are uni-directional, but many are multi-directional. I think most smart phones have them inside. That's why your phone lights up when you pick it up. That being the case, many are built with the assumption that they will not be connected to any external device during measurement. They can be very small, very light, self-contained, and relatively inexpensive. I've seen them used in high school robotics competitions.

You now know everything I know about accelerometers. I suggest a google search and then several direct person to person telephone conversations with application engineers at various manufacturers. (No email.)

 

[MintJulep]

The "force profile" of a load cell moving through a chute is rather likely to be different from the force profile of your product moving through the chute.

Someone recommended that I use a load cell

That was a bad recommendation. Say "Thank you, I've researched that and determined that it won't produce any useful information."

I will try to be as specific as I can.

Try harder.

 

[MikeHalloran]

Further to MintJulep's point, you already know that the material's yield point is being exceeded somewhere in its journey. A time profile will add nothing useful, but it will cost money and time.

Your time would be better spent watching the machines, and especially talking to the operators, jobsetters, and toolmakers involved in keeping the machines running. There's a good chance that most of the people on the production floor already know how the laminations are being distorted, and some of them know how to fix the problem. ... and a few of them may be right.

Spend a few days out in the factory, chatting up the people closest to the problem.
DO NOT carry a clipboard; that will frighten them.
DO go to a quiet place after each conversation, and record what you found out.
Include the facts.
Do not include your opinion.

If a direct worker actually gets close to the solution you use,
find a way to reward that worker.



Mike Halloran
Pembroke Pines, FL, USA

 

[rb1957]

You say this part has been in production for awhile. Why are parts deforming now ?

It's hard (for me) to see where the deformations are happening. In the 2nd pic the parts look to be somewhat controlled by the guide rod. Has the chute (in the 1st pic) worn so the chute is no longer tight on the part ? or worn so there's friction (that wasn't there when new) ?

How many of these things are getting deformed ? Presumably there's no rework ... it looks like you're making a bejesus lot of them so their value is about nil (so why rework). Does the deformation cause a problem with the next stage of production ?

another day in paradise, or is paradise one day closer ?

 

[MintJulep]

Step 1: Understand the problem.
Step 2: Solve the problem.

Skipping step 1 very rarely results in the desired outcome.

That is why if I was able to give a definitive answer or measurement as to when the bend in the parts is occurring, I would more easily be able to find a solution to the problem.. but it is hard enough trying to find the cause of the problem in itself.

Good, you understand that step 1 is important.

It could be that the part diameter is too large as compared to the rest of the chute diameter, but there is no way for me to give a definitive answer on what is causing them.

Getting that definitive answer is step 1.

a tight interference fit with the part and the chute is in my opinion the most likely cause for the bent laminations.

How have you conclusively demonstrated this? Or is it just a guess? Again, see step 1.

 

[ak369]

The company so far, year to date, has produced about 145,000 dollars of scrap just in bent laminations.
It is the highest contributing scrap factor in the company.
Essentially, that would be over 300,000$ lost by the company just in bent laminations occurring every year.
With the first picture, I went out to the press, and to me it didn't look like there was much room between the parts and the press. Just as a general observation, the parts (especially the larger ones) are tight on the inside of the chutes. However, I don't know if it would be feasible for the company to go and replace every chute where that is happening.
The chutes are pretty worn out and old, there seems to be a lot of friction between the parts and the chute just based off of my observations. In my opinion, there should be a larger gap to allow the parts to pass through more freely. Perhaps opening up the top of the chute would help.
Finally, once the deformation is observed, the press operator stops the process and tries to stop the deformation from reoccurring. The parts are only sent to the next stage of the process once the issues are corrected.

 

[ak369]

@mint julep
I'll just go and measure the diameter of the parts versus the chute dimensions at various points. You make a good point although kind of harsh it's true. I'll let you know what I find.

 

[rb1957]

There are many process quality programs/methodologies that are designed to solve issues like this, eg (and no fan) 5S, CQI, ...

as MJ says understanding the root cause allows you to fix the problem.

Someone made this machine and presumably it worked. Now you're a scrap metal manufacturer (clearly not good). Now you interrupt your production to fix issues. There's a clue there ... what does the operator do to "clear" the problem ? what "fix" works in the short term (until the problem recurs) ? This may lead you to the root cause (or not !?)

Maybe take video of the machine working and see deformed parts being created ?

I don't envy the position you're in, as an intern/ new grad, being asked to solve a production issue like this without much help ("help" is probably someone asking "why isn't it fixed yet?").

another day in paradise, or is paradise one day closer ?

 

[IRstuff]

Seems to me that the system was initially designed for a very low volume production. The guide cannot possibly support that much load and not cause the laminations to drag along the bottom of the chute. This would cause the laminations to drag along the bottom of the chute.

Where exactly on the laminations are the bends? Is it random or consistent? After this much time in this thread, I've not seen any pictures of the damaged parts. Seeing the actual damage might tell you volumes.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[TimSchrader2]

Hello

Can you duplicate the problem?

The above theory about the wieght of the parts and the rod sounds like worth a go. Has the quanity of parts made increased so more wieght on the rods? Especially when going down the turn. Can use see marks on the chute where deformation occured?
What has changed?

 

[btrueblood]

Grab an Iphone, it's capable of recording video at a fairly high rate. Or go buy a panasonic point-and-shoot camera, ditto the high frame rate. Record some video until the parts come out defective, then study the video to see if you can see the deformation occurring.

Talk to the operators, what are they doing to correct the deformations? How are the deformations found and where? Can they be detected at an earlier point in the process, and could it be done automatically, or could the correction be done automatically, or both?

Pictures of the deformed parts, and maybe some arrows on the pictures telling us which direction the parts are being moved and what process they are emerging from/going to.

 

[ak369]

@btrueblood

Okay, I will do that. But they occur inconsistently, I hope that I can find one when it happens.

 

[btrueblood]

What do you risk by being "shot down" vs. what's to gain if you fix the problem for them. Ya gotta try, right? Maybe finding the right person to approach, and mentioning the scrap costs...somebody in charge of money, say...

 

[ak369]

@btrueblood
You're right. But there are people who have worked in this company for over 40 years. It's just my second month* at the company, that's why I'm not sure how much my opinion matters.

 

[handleman]

ak369,
Just a word of warning... I would recommend confirming with management prior to posting so many photos of your machinery and production parts. You may think nothing of it, but these may be trade secrets or other proprietary information that should not be shared outside your company. At my last company (which, coincidentally, also made stators for HD, although different ones), the pics you posted would have been a HUGE no-no.

 

[rb1957]

apologies (I'm told I can be blunt) but right now in your career your opinion doesn't count (for much).
what experience do you have that supports your opinion ?

don't get me wrong (which I admit is very easy), I'm not saying don't have opinions; just don't expect other people will value them.

has someone given you this problem to solve, or do you think it is a problem that needs to be solved ? It could well be that people don't think scrapage (and interrupting the line) is a big problem. Maybe it isn't, maybe the same people don't know the magnitude of the problem ($s in scrap, numbers of lost production, ...) and would have a different opinion if they did.

If someone has given you this as a problem to solve, then someone thinks it is a significant problem. Maybe they're looking to see how you approach problem-solving ?

Collect data, Observe differences in the data (some machines behave well, others deform parts), collect more data, ask questions (why would machine A work differently to machine B ? ...). Get some CQI training (6 sigma), or at least read (and teach yourself) about it. The key to fixing problems is understanding the root cause.



another day in paradise, or is paradise one day closer ?

 

[MintJulep]

Fundamental approach:

Understand if defects occur to single parts or to groups of parts.
Look at the defects that occur and understand the forces that would be required to create the defects.
Look at every inch of the process and understand if it is possible for the forces that cause defects to occur at each inch.
When you identify places where it is possible, understand if it actually happens.
For locations where defects happen, understand why.
Design and validate solution.