QC Verification

• Weigh it empty.
• Fill it up with water and weigh it again.
• Subtract the smaller from the larger to obtain the weight of the water inside it.
• Divide by the density of water to obtain volume.
• Divide by the diameter squared and multiply by 4 divided by pi.
• The answer is the length.

Thanks for the reply, but if you put water in a coiled hose you will have to take the strapping off and straighten out the hose to get the water back out. We were hoping to be able to find the length without taking the finished product apart. Again though thanks!

<...you...>?

Invitation declined.

You dont even need water on it. Cut 1m of the corresponding size of hose, may be 20 times, the more times the better, weigh each one, then find the average weight to 1m ratio. Now use that ratio to find the length. Say you have 0.23kg/m average and you have 5kg of hose, you would do calculation (5kg hose/0.23kg/m). A good quality hose will aproach average 0.23kg/m value. Otherwise, you spend some excess material on the making, or not too much material for the making which would turn out a substandard hose.

If you want to be more accurate, make the cut say at 5mm sample or less for corresponding size of hose

My buddy kulas and james said this will work. Homework done, grab a bud, now then.

Have you tried the techniques used to measure belt in a roll:

http://www.handymath.com/cgi-bin/rollen.cgi?submit=Entry

I am not sure how accurate it would be for hoses but you could run a few experiments and see if it is a valid course of action to develop.

Depending on how your encoder measurement is set up, this will likely be your most accurate measurement. You should be able to perform an in-situ weighing of the hose (without water ) to look for gross errors. Any volume measurement technique will require spooling out your product to verify verify the walls didn't collapse, bulge or deform in any other fashion. You might as well get the tape measure out.

At one place I worked, they had calibration stickers on all of the office rulers. What they used to calibrate them, I have no idea. I also thought of what they would do if a ruler was found to be out of calibration. I suspect that someone in QA misinterpreted the ISO rules.

Here we go again. One cannot "calibrate" a ruler, only "verify" that it is accurate. Calibration means that you can adjust something to a known standard. Unless they were very expensive rulers with adjustable components,,, well,,,, . I do use a steel rule to verify some of my measuring equipment, and it is verified every year to verified and traceable gauge blocks or other such equipment.

So why can't the OP verify that these blinking hoses are the right length?

Hurrah for testing, then.

Makes a lot of sense; thanks, I'll let the QC department know.

That's a good answer North. Weigh the finished apparatus, take out the constants that is not hose. All that is is left is hose at inches/pound or whatever units you select. BTW is it starting to freeze up there yet or are we still suffering from global warming?

Oh ya... we are cold. I am having to warm the car up in the morning to keep the Mrs. happy. Have had an inch or two of the fluffy stuff. A few communities have already enjoyed their first blizzard.

Oh joy oh bliss...

Technically it can be done in many ways, but methods like the one you suggest are subject to many factors, controllable like temperature and virtually uncontrollable like strapping consistency etc, so getting good accuracy and repeatability won't be cheap. Now since you have access to both ends you can have a custom device manufactured to send a short acoustic burst in the one end and measure the delay for the sound to reach the other end. No rocket science, plus airborne sound can easily be distinguished from the one propagating (much faster) in the hose material, and with some calibration and temperature compensation you can go to less than 5% error. If you want better accuracy, rather try and improve your winding system to make it more fool proof resistant. S.M.

You do have a bit of a dilemma here.

I used to work in the rubber manufacturing business and we made hoses of all different lengths so here are a few things that come to mind.

• Rubber or thermoplastic hose varies in weight per foot but probably not enough to make a large length error? You would need to find an average weight per 100 feet.
• I wouldn't try air or water capacity. Too difficult to get an accurate measurement and then you have all that water in your work area!
• Your best bet is to calibrate your encoder, do a test run and verify the accuracy of the encoder calibration.
• Check 1 out of every 100 pieces and you should have a good production run!

Good luck!

I agree, I'll give it a try. Thanks!

Do you really understand the word " compressible". If not please measure the air.

I never its not doable, i say its preferrably you will introduce lots of error.

Gases are compressible, in reality, it would not be that accurate....as someone already mentioned, liquids would be more accurate and a good blast of high pressure air should clear any residue for weighing.

Weighing the hose before and after the test would tell you if liquid had been left inside or not!! Just add the difference for simplicities sake to the amount collected, then you can blast out the rest and not need to catch it!!

If you only need to test 1 in 100 for example, surely something like that could be helpful...?

You could use air to pump the insides up to, say, 100KPa and then use a large "syringe" to add a known volume of air and measure the pressure difference. Work it out from there.

Practicalities; getting an accurate "syringe". Perhaps an old hydraulic ram. A plug at one end can have a gauge fitted.

This will give a volume but expansion of the hose, uneven sized bore and leaking plugs at the ends will affect the result.

A known length can be tested in this way to establish a 'go - no go' standard.

I think weighing it will be easier.

Jim

A more or less unusual way would be to build up a pneumatic potentiometer :

pressure supply regulated (low pressure) + one fixed orifice + a T junction with a manometer + the hose you want to check.

If you calibrate the device and if production is consistent you could get the length with quite a high accuracy.

Why not weigh it?

Blow a pithball and line through it, mark the line and then pull/blow it out and measure the line. That's how to measure length in a conduit! It should work for a coiled hose too.

That should certainly work with a rigid wall hose or pipe.

You seem to be new at Quality Control/Quality Assurance and verification procedures.

The first thing you need to do is determine if your hose cutter is consistent in its measuring. This means you need to verify/calibrate the encoder you are using on the line.

After that, you need to do 'lot testing' of the produced hoses. You take a random hose from a produced batch, cut the straps, lay it out, and measure it to confirm it is the length claimed, within the tolerance for the product line. The tolerance is typically set by a collaboration between Production and QC, for the sake of argument, let's say that a '100 foot hose' is defined as 100 feed plus or minus one half inch, so any hose 99'11.5" to 100'0.5' is 'good.'

Now, if your plant is doing "100% verification" then EVERY hose needs to be measured individually to confirm its length. If the plant were doing that, however, the time and cost involved would be rolled into the price of the hose, making it much more expensive.

Now, if you've got the floorspace in the production line, you could have a second encoder after the cutter to measure the hose, and if it was 'out of tolerance' have it diverted to a 'reject bin.' but that's additional space and expense that the company may not be willing to invest.

Good luck with your new QC position, by the way.

You should be able to set up a jig to cut lengths the same. Then periodically check the hoses by random sampling to make sure that your jig system is still in calibration. You should also put a tolerance of +/- in the overall length, so you're covered. Like Subway should have done to prevent being sued because someone was served a sammich that was only 11-1/2" long instead of a foot long.

Furthermore, you have the weight of the hose at 100'. You can just weight the hoses and have set tolerances for weight.

I can understand material expanding and shrinking due to ambient temperatures. So store the hoses in a room with a constant temperature, so that linear measurements are consistent.

You are asking for a "Quality Control" answer to a process that could have a "Quality Assurance" outcome.

Calibration of the encoder is the important step and then adding some means to verify the distance from the start of the hose to the end.

Polythene pipe here is branded by the manufacturer with batch details and such along the length and a marker every m. By encoding the length onto the hose, you can see whether it is the right length (Branding equipment coalibration to be in place) and for users of the hose, they can calculate how much is left on the roll once pieces have been cut off for use.

So, brand a marker every m and then the QA needs to verify the printer calibration and the QC guy only needs to measure the amount at each end to the nearest m marker on the hose.