Neuro Hydrophilic Coating Machine

Project Highlights

Neuro guidewires are coated to make them lubricous so they slide easily when inside an artery.

This project involved neurological guideswires requiring a hyrophilic coating on the front section of the wire. See image below.

The wires were being coated by an existing, homegrown process.

This three step coating process is a time-consuming batch process. It uses expensive, nearly unmanageable coating chemicals which the customer insisted on:
  • Primer coat
  • Base coat
  • Top coat
The base coat was very difficult to work with:
  • Would stick to anything it touched.
  • Dried like concrete and was nearly impossible to clean up. Almost anything it touched other than glass or PTFE was considered a throw-away consumable because it could not be cleaned at all or cost effectively.
  • It had to be used quickly before exposure to air ruined it.
In addition, the existing method isn't economical in its use of the chemicals.

How the wire is coated

The existing machine which worked well was designed to apply the hydrophilic coating part to the exposed wire by drawing the lower 12" of it vertically upwards thru a hole in the bottom of the cup and then thru the liquid coating surface (cup is lowered relative to fixed hanging wire). One cup is used for each coating then discarded as it's not not cost effective to clean.

Fifteen cups had to be quickly filled to 2/3 full by hand with a squeeze bottle.

Labor intensive,

...messy and very wasteful.

The Goal

The new process aimed to reduce the consumption of chemicals used per wire, eliminate or drastically reduce consumables, and reduce the labor time to prep and clean the machine.

The Team

I was project manager / design lead and assigned a cross-functional team of four.

The work

Analyzing the problem

The existing process's achilles heel was the labor and high cost of consumables to handle them. Analysis revealed:

  • Time to fill the cups was a major road block; it was slow.
  • Cups were being over-filled by the squeeze metering bottle.
    • Over-filling created a troublesome mess requiring clean-up.
  • Operators over-filled to avoid the cup running out of coating liquid.

Design concept

After some inital brainstorming, I decided to modify the existing process rather than replace it.

The process modification I conceived of was to deliver each chemical directly from the 55 gallon drum to the point of application - "the cup".

I envisioned that at the cup point of application, automatic dispensing heads would meter out exactly the right amount of liquid into 15 cups simultaneously when required.

This sounds innocently simple, but when considering the chemicals involved, it was a very ambitious design solution indeed.

The dangling carrot was too hard to resist

The potential cost savings were huge.

  • Labor time to fill cups would be eliminated
  • Wasted coating would be signicantly reduced

I had taken on very challenging projects before (see Profile grinder project) and I was up for the challenge. I strategically planned a design approach which factored in two possible design solutions to ensure success (the second being less risky). Based on my track record of success and good business sense and judgement, I was given the green light to proceed.

Building the Prototype

The design engineeering phase began which involved:

  • Me converting conceptual sketches into AutoCAD
  • Software programming

A test rig incorporating one pinch valve and hose from a simulated parallel setup was put together to test the automatic cup fill concept.

Proof of concept test - 1 cup fill

The simulated one cup test using water (same viscosity as base coat) worked well after the system was tuned to dispense the correct fluid level into the cup. We committed to build parts required to do a 15 cup fill test.

Proof of concept test - 15 cup fill

The dispensing consistency across all 15 cups wasn't there. They were all different levels. We were puzzled. We scratched our heads and spent time troubleshooting the problem. Finally, the penny dropped.

When the tube pinch valves released, the crushed hose spring back was inconsistent across fifteen stations, so pressurized liquid dispensed through orifices of varying sizes. This situation didn't arise on a one station test.

This would not be easily solved, however failure was not an option.

We were so close to an efficient solution...

To solve this I used my engineering knowledge of materials. We needed tubing which had a wall that could bounce back quickly after pinching. I ordered tube samples and ran a lot of tests.

We made significant iterative improvements with soft wall tubing but the dispensing accuracy still didn't meet our requirements.

Then I had a sudden inspiration: thin wall PTFE heat shrink tubing. It recovered perfectly as the fluid pressurized it internally.

The problem was that with typical off-the-shelf shrink tubing, unshrunk ID's are not manufactured to tight tolerances because it's typically not necessary as it gets shrunk down to the size required. More accurate tubing could be manufactured but, it would have been a special order with a 6 week lead time. We were out of time.

Backfall position envoked

We had to switch to the backup plan; it wasn't easy when we were were so close, but it was a business decision I had to make.

Fortunately, my second backup design idea deliberately utilized a lot of the same new machine modifications. It was influenced by expertise with glue dispensing systems for guidewire assembly. The idea was to use 15 resealeable pipettes which would be placed in permanent holders on the machine as shown in the photograph. Each pipette would be filled with sufficient coating liquid to coat ten wires. Once the pipette was empty, it would be discarded. So, instead of a continuous flow of coating from the drum, a localized reservoir of liquid would be utilized.

A glue dispenser (reworked for dispensing hydrophilic coating) would feed clean, pressurized air into the top of pipette above the liquid through a cap. The air would displace coating fluid out into the cup, instead of the old system of pinching off a continuous, pressurized coating feed tube. When the correct amount was metered out, the air would be shut off and a small vacuum applied to halt the flow and prevent drips.

Photograph shows 15 pipettes aligned with cups on a carriage driven by a ballscrew. As the filled cups descend vertically, they coat the affixed guidewire.

Pipettes move away from cup on linear slides after filling to allow room for wire to pass vertically through cup.

Revised Prototype Testing

Success at last!

Success

The revised coating machine was effective and efficent and drastically reduced operating costs.

  • Simultaneous automatic metering of fluid into the cups reduced the cycle time.
  • Automatic metering of coating reduced coating wastage due to over-filling and spills.

What I Learned

This had been a tough project, but thankfully my engineering design expertise allowed me to perform course corrections when necessary. In addition, I learned that technological advantages provide a competitive edge and need to be encouraged but managed carefully. Keeping an eye on R&D costs must be balanced against potential cost savings and need to be implemented in a timely manner.