Jun 02 2007

Got Water?

Published by at 5:28 pm under Uncategorized

Ever wonder if your plant is supplying enough water to the machines and molds? Here’s the reasoning:

This is always kinda fun:  Ever have flowers in your house you barely water?  I even had a friend once who killed her cactus plants because she thought desert plants didn’t need water.

OK, let’s relate it to Injection Molding.  It’s almost a no brainer to standardize as much as you can on the shop floor, but sometimes you shouldn’t.  I’ve been in a few shops that have in large white markers written on the press platens “ENGLISH BOLTS ONLY” or “METRIC BOLTS ONLY.” Before this practice many technicians cross threaded bolts.  The frequency of thread inserts in the platens was a sad commentary to this practice.

I have a client who started off literally as a small company.  For a while he was the king of the 50 ton machines.  He lived and died with small machines, small parts, and small molds.  Naturally everything was fitted with ¼ inch water fittings:  the molds, the machine manifolds – when they existed, and the temperature controllers.  The only large fittings were the ones connecting the temperature controllers to the “tower”.  Fine.  Right sizing for the right type of molds and machines.

But as many molders have had happen, the curse of success got him into bigger and bigger machines.  Larger machines meant he could charge a higher hourly rate and use more material.  Naturally the overhead cost went down as profits went up.  As local manufacturers would move their operations off shore, he’d step in at the auction and get his additional machines at Garage Sale prices.  It was then a simple matter of adapting them into his operation.

There was a silly problem to this transition – waterline sizes.  He had hundreds of straight, 45 and 90 degree of ¼ inch male and female waterline connectors all connected to the appropriate hose diameter in varying lengths (short, medium and long).  His setup guides not only showed the waterline pattern but even instructed the length of the hose.  Nifty!

When faced with the problem that larger machines wanted more GPM going through their molds, he took the path of least resistance, ignored the problem, and retrofitted everything down to his system of ¼ inch cooling lines and connectors.

When measuring the proper flow you look for temperature exchange and a rough approximation of turbulent flow.  Small diameter cooling systems require less GPM or LPM than large diameter cooling systems to get turbulent flow.  Your definition of turbulent flow is through the largest diameter, regardless of where you measure it.  A crosscheck of the heat exchange capability of the system is to measure the inlet and outlet temperatures.

The rule of thumb (because I hate the mumbo jumbo of Reynold’s Numbers) is 1.5 GPM will always cause turbulent flow.  No, not in a two inch diameter circuit and it is a bit much for very small water systems but it’s a good gross measure.  More importantly is the temperature difference.  A 3 degree F (1.6 degree C) or less differential between inlet and outlet is a good indication of adequate cooling. A 5 degree F (3 degree C) differential shows marginally even cooling, and usually any differential greater shows your cooling is inadequate.  Does bad (inadequate) cooling mean you make bad parts?  Nope, you just slow the cycle down until you can pull enough heat out of the plastic before you eject it.

When he was on vacation I did a test at this facility with a friend who worked there.  I put two size adaptors on my handy temperature-pressure-flow meter and hooked it into one of his molds on one of his 450 ton machines. For a point of reference of heat exchange I measured the ejection temperature of newly ejected parts. The GPM was embarrassingly low, but even worse the temperature differential was 13 Degrees F.  Ouch! Just for giggles we (the engineering manager and I) put a booster pump on the temperature controller, brought the pressure up to an obscenely high pressure but maintained the temperature settings on the controller.  I won’t even tell you about the flow, but the temperature differential admirably dropped below 2 degrees F.  I took out my IR Pyrometer and began checking the ejection temperature of the parts again.  Yup, as expected, they were noticeably cooler.  We cut back the cooling time until we achieved the same ejection temperature as we had before super-charging the water circuits.

The next result of improving cooling through flow was cut 5 seconds off the cycle time achieving the same quality parts.  We then unhooked the booster pump thankfully not having burst a waterline or had a connector come loose and whip us to death; the ejection temperature of the parts began to rise until we increased the cooling time to its original setting.

Before we went to lunch I looked up the price of 3/8 inch male and female fittings appropriate to the larger machines, hose, hose clamps etc.  We took a quick guess at the number of molds, fittings for each, and the spendy labor it would cost for the transition.  Between the time the waitress took our order and the food was plopped down on the table we did some serious keyboard poking on a PDA.  The total cost was few thousand dollars.  We then did a productivity calculation to see if we’d save anything.  ROI? Four retrofitted machines with larger fittings would free up almost a half a machine’s capacity in 12 months.  Then I asked what we consultants call the “Duh!” question: ‘Why not implement it?’

Resistance to change usually comes from the Generic Excuses (1) “We always did it that way, or (2) What we’re doing is convenient.”  Sure enough this wasn’t a novel idea, my friend had brought it up before, only to get the Generic Excuses saying it was going to be too difficult to have two sets of connectors and hoses and such.  Since production wasn’t in his job description, he’d brought it up once, justified it, made the presentation to the production manager and the company owner, and then let the matter drop.

This was an interesting exercise in Lost Opportunity:  The company was perfectly happy walking away from a huge profit because they’d initially settled on one diameter waterline.

Go buy a flow meter and check the inlet and outlet temperatures.  If you see a huge discrepancy you’re under-watered.  If you increase the pressure to increase the flow to reduce the temperature difference you are only overworking your pumps.  A mold can only give up heat at a certain rate, but if you can pull the heat out faster than the mold can give it to you, you are doing the best you can with the mold’s cooling system.  Lowering the temperature doesn’t do it. Its volume through the mold that gives you heat exchange.

Is the study and a retrofit worth the money? If the ROI is there, take it.  But be aware of the emotional factor of resistance to change.  Too many people mistakenly view change as flying in the face of LEAN.  To them cutting back is Lean; adding to, isn’t.  When you can explain this logic to me in small easy to understand words that don’t have the phrases like ‘value chain mapping’ in them, please e-mail me.

* * * * *

This article is virtual.  If it scares folks, delete it.  But if your ego can take it, let middle management take credit for something that contributes to the bottom line.  This time it will be hard dollars not PowerPoint slides.  It might be worth it.  If not; do the studies, present your findings, and file the results in the “I already told you” file but don’t put a date on it.  It should be periodically dusted off and re-presented when the Unending-Search-For-More-Profit-(with some Japanese named philosophy, complete with Karate Belt type certifications)-Ninja-Geeks come howling at your door.

Your choice.

One response so far

One Response to “Got Water?”

  1. Rick Hagforson 15 Jul 2008 at 4:02 am

    One of the most overlooked pressure loss in a plant comes from the feed throat cooling lines. They are probably wide open and could be throttled back without significant change to material flow. Especially if it’s Engineering Grade Resins.

Trackback URI | Comments RSS

Leave a Reply