Worsted Combing - Rise of PC Computing

How the arrival of PC computing assisted innovation in worsted combing?

Its best first to explain that worsted comb is part of pipe line of fibre ( in my work wool ) to the finest suit a minister might wear. Fig1 is a picture of a comb. It is a complex mechanism placing large number of clocks or mechanical robots to shame. Each machine costs in order of $250,000 and has a useful economic life of 8 years before it performance degrades. And in worsted combing performance is everything. Large scale worsted operation would use about 60-100 combs.

The comb formes a bottle neck in production and costing of the combing plant.

How is this related to computing and 3D printing.

In 1988 at dawn of the age of PC by Microsoft I was tasked by CSIRO to improve worsted combing. Combing is a process of removing short fibre and allowing theoretical limit of 31 fibres to be used to make very fine and executive suites from wool. Essentially machine design of comb stayed the same until 1995.

Here is some Background.

Combing originally done by hand was automated first in England and than France at dawn of Industrial revolution. English originally lead the technology which they attempted to monopolized on pain of death. England's traditional enemy - France responded by building the worsted combing based on linear motion rather then circular motion. Subsequent technologies were based on the French design.

For more information look up wiki on combing https://en.wikipedia.org/wiki/Combing.
Australian Wool clip is within range of fibre diameter ( 12-22micron) that is worsted combed. Same type of machine can comb fibre from flax to cotton.

In 1989 computing power was limited and this particular comb is driven by 10 cams. Each cam is a complex shape almost reflecting "hand writing automats" of the French Swiss fame in late 18 century - you can think of a cam as a program required to be executed.

Here is a figure depicting only one motion synthesis of a single mechanism.

The blue line actually depict the motion that occurs at 3.5Hz. Each Green dot indicates an adjustment to a motion. In 1989 it took up most of the memory to draw this in an early version of a PC CAD system. This was only one of eight sub systems required to remove short fibre and vegetable contamination from the entire Marino wool clip.

Here is how precise the motion of the pins needs to be. The scale is in mm, and cost or removing each contaminated fibre was $1. The efficiency of the comb was in order of 95-99%. This was not helpful as laws of probability greatly fluctuate at that order of efficiency. It influenced how well an experiment can be performed and return a valid result. Complications were also compounded by the fact that contaminates had to be measured and sorted by hand.

However at time it was agreed that if we could double the production and reduce maintenance the cost for a combing mill will fall.

Natural approach was to speed the comb up and deal with issues of fibre breakage. Sounds easy enough...?

Here we look at dawn of the PC computing, question was how to feed all this information into limied computer memory of the 1988. As result of my work I was able to divide the computing requirement on the mechanics into 6 parts. Four dealt with machine mechanical elements , 2 dealt with fibre model and others dealt with theoretical aspects of fibre breakage and noil ( waste removal ) at efficiency of 95% and above. The math and simulation needed to have no error.

Building a Computer model - CAD only. So here is only the apron motion (1 off 8 mechanisms ) CAD solution.

The image can be compared with picture of the actual comb. Light blue lines indicate gears.. Here is same figure in detail.

Simulation was written directly into the computer code using Lisp and C++. Access to Adams software was limited and at my CSIRO pay grade not warranted. So in best of CSIRO tradition I wrote my own simulation from low level code.

Simulation Figure shows the adjustment of the apron. Apron is a mechanism that delivers overlap on each combed tuft of fibre. Imagine each tuft like a sheet coming from the photocopier, but in case of comb each tuft needs to be overlapped by a percentage - much like the tiles on the roof. This mechanism controls overlap and timing in synchronization with other motions at 3.5Hz.

Simulation is critical to the insertion of the "top comb" or single line of pins described earlier in impeding the flow of contamination.

Each Sim block is a computer program written in C/C++ and connected by a variable path. One can appreciate that motion is controlled by cams. One of the cams was my break through design I will talk about physics of its performance in another publication. Sim or Simulation was required to assure the experiments were performed with fibre following exact same path as used in normal commercial comb.

Since a comb was controlled by a set a program controlled by some 10 cams , I also needed to create a technology to make my own cams. At that time CSIRO division of wool technology owned a CNC machine capable only of 64K memory - hence unable to make anything more sophisticated then a CSIRO badge carved from aluminium and usually given as mementos to visiting Ministers of the Government.

I created a CNC machine program that would make Cams from data using C++, then a new language. In process of modifying the existing machine I also created a computer program to digitize the surface of the cam. Since my back ground was in signal analysis I used Fourier to clean the digitized surface. ( Later on Fourier was replaced by Wavelets as preferred analysis tool ). This was important as cams operate on derivatives - velocity, acceleration and impact - 3rd order. In numerical terms its hard to obtain higher order derivatives - Fourier made that easy and Wavelets improved the result.

This is a 1990 picture of our digitizing table for cams and spring elements. Note the IBM-AT box running Win 95. That particular computer had 8 upgrades in Hardware starting with inter 386/7 and Win 3.1. Its still in use.

This is the same year picture of a manufacturing technology developed at Div of Wool Technology ( graphics card level interface, CAD and low level communication over RS232 there was no USB 1.0 only LPT or RS232 ). Our CNC technology operated on 300mm sq table.

Here is picture of my technician operating DM4400 CNC machine.. The software was capable of dynamically interacting with the operator by choosing the tool from a rack of 6 and positioning the cam for optimum performance ( translate and rotate ). It was all done with minimum documentation from the manufacturer of the CNC in Taiwan ( make do principal operated at our non-management pay grade ).

In 1995 CSIRO cut funding to our project and Dr.Kirby and myself had to leave - just as we have solved the problem and just as we were getting tools and IP in place. After leaving CSIRO, computers improved and have improved since by 100 fold. at about 2000 I was able to CAD comb as a singe drawing 5 years later I was able to make it 3D and fully simulate it in ADAMS and my software.

In 1995, I formed a company called "Comb R&D P/L" that run aground immediately with legal problems as CSIRO demanded that I do not use any technology I developed at CSIRO. There was also a question on a provisional patent and IP I generated as CSIRO required to exploit the outcomes.

This was followed by a full construction of a 1/2 scale comb as it is as a demonstration my ability to CAD and CAM every part of the technology.

This is the apron drive side as seen in the original photo.

This is main drive for a cam system..

This is control cam shaft as seen from the front of the worsted comb. with a clamping mechanism removed.

The Model of the comb 1/2 scale will be a permanent loan as an exhibit at the National Wool Museum. With its associated equipment used in various developments. Here is a clamping mechanism as used in my Masters degree on fibre friction also destined to be a Museum exhibit.

The system operated at 240 cycles per minute in synchronized motion on an area less then two postage stamps and width of 380mm the width of the comb apron.

This work is still current as technology of commercial combing has not change.

Here is a picture of a Combing Mill in Australia one of many that have shut down since.

Just keep in mind that each one of those machines is critical according to Australian Bureau of Agricultural and Resource Economics and Sciences ... ABARES is a research bureau within the Department of Agriculture.

My company still posses technology to double the productivity of worsted combing and that technology has steadily improved. All current worsted combs in world made by NSC carry components of development work done in Geelong - CSIRO division of wool technology before it was closed.

My future posts on this topic will include description of fibre performance in the comb, fibre friction, details of what was given away and how it impacts on technology other then that used in textile industry.

Next post, depending on the interest will described new type of cams that facilitated CSIRO taking out provisional patent after NSC France informed us in 1989 that "anything that CSIRO would like to do should be passed on to their engineers to see if it can be done or was attempted before", I grew up on streets and I was not that gullible with IP- so we ( Kirby and I of CSIRO ) developed critical component not seen before with aid of Computer technology CAM/CAD/CNC and now 3D printing.

I expect that our version of cams will make it into motor industry as it is natural that textile machine developments were used in dawn of Automotive industry.

My next post will be talking about a non-traditional way of designing cams and evaluating cam efficiency ( its is more suitable to signal analysis then traditional mechanical engineering ...)

This is how worsted combing has benefited from rise of computing and associated technology. We acquired 3D additive technology and have added its value to worsted combing developments.