#xcsiro ToD21thCentury 221-230: Australian Worsted Comb Story
Dr.Nick.Sokolov CRD
PhD(Wavelets).MS.BE.(ME,EE), Owner CRD P/L, Father of the Australian Worsted Comb Fibre processing Technology
=====>>> This is my 100th publication ....
....on a LinkedIn showcasing the references to my knowledge and missed opportunity by CSIRO. This publication is a collection of References I used in my everyday research, and experiment building between 1985-1995. Currently I am still doing the same work with advancements in Australian made Worsted comb - capable of outperforming the NSC Schlumberger.
enjoy reading notes on Electrical and Mechanical engineering, numerical method of combing and Wavelets based on Shape derived functions as filters ...
This is all useful stuff. I hope it makes you choose a suitable path in life.
8th=eighth (8th/8) of the "Source Books" and Simulink in MATLAB.
In 1989 I was writing all my analytical , digitizing and CNC code. It was an exercise keeping track of various projects.
This was just one portion of programming back then -a numerical method to solve for the noil (waste product of combing) &tuft (overlayed shape producing a combed sliver)
Each numerical analysis has its problems - so there are no real books that one can use for this particular problem. One can see that noil generated in subsequent combing of the fibre population is almost equivalent to an error in some numerical integration methods. I needed a integration method that had zero rounding error when compared to the original population distribution p(l).
To calculate the tuft shape ( combed sliver segment ) or the noil ( wastage or removed undesired fiber length ). I had to handle a triple integral. Related to following.
1. Position of the fibre ends in the uncombed sliver
2. Position along the individual fibre length
3. Change in the fibre distribution of fibre length after circular combing
4. Change in the fibre distribution after front comb combing.
5. Ratio of Stationary to moving fibres.
There were 2 competing theories:
1. Create a numerical model for 1968 Belin & Verhagan theory
2. Create a numerical model for 1991 Kirby & Sokolov model
AIM: Create a proof that it was valid numerical model
This all had to be done with zero error for Noil generated.
Ref for more detail: https://lnkd.in/gc4Rt5A6
Originally I was working with Dr.Barry Harrowfield on Belin model, in 1988 and within 3 months I needed to write a graphic interface and discover short combings in a numerical model. He did not understand what were my problems or he failed to understand the complexity. We had a falling out - he appointed Mr.Ken Atkinson as a project leader.
Once Dr.Kirby and I settled our differences we produced Kirby-Sokolov model for noil &tuft
This did not go down well with Berry and Ken. At one stage Dr.Barry Harrowfield called Dr.Kirby's Model a circular argument and walked out. Disregarding my numerical solution. Followed by Ken who said: "I am with Barry"
==>>What is my point in all of this?
As 1990's progressed more useful computing tools appeared -such as MathConnex, VisiSim and finally more useful version of Simulink. A graphic driven blocks of C++ code &older programs that can reduce complexity of keeping track and moving data about. BUT by that stage we were removed from combing project -they had my patent and IP and Kirby was struggling to find another job.
==>> Read More https://lnkd.in/g3zVeiG4
Wavelets: continuous, discontinuous &now pseudo (as defined by my work)
==>part of my 100th article/Publication for Linkedin (draft)
7th=seventh (7th/8) of the "Source Books" and Wavelet Toolbox in MATLAB.
===>>>Why was signal analysis in cams important??
In 1993 I became interested in adding sensors to the comb. There were traditional sensors such as positioning, load and rotary encoder. This was not enough -sensors only indicated what the position of the machine was but no dynamic fibre behaviour.
===>> What type of the signal was required?
Its simple - I needed to filter out a behaviour generated by the cams. With Sokolov Cams that was possible as - dynamic effects produced by the retaining springs were removed.
===>> Why the filtering was not done with Fourier FFT type analysis?
That was obvious at least to me. A Cam shape would require many frequencies injected into the spectrum. For instance an eccentric circle - the simplest of all cams would have a single peak adding a 2nd frequency produce a "peanut" shape .
I even toyed with an idea of multiple frequencies to get the desired cam shape ( but CSIRO made us redundant )
===>> Why did I end up using Wavelets?
By definition of a Wavelet I was inspired that I can turn CAMs into a Wavelets &effectively measure the acceleration of the comb or speed variation in its operation.
That was useful BUT what was more useful was to filter out the CAM shape as a Wavelet and leave behind the "signal of interest".
===>>>WHY?
In a well behaved machine -signal generated is not mystery- it is what it is a predictable system. Sokolov Cams with inert dynamic behaviour allowed that type of analysis.
If you look at the JPG - there is a yellow tape stack on the surface of the CAM - Sokolov Cams make the spring inert and the motion detects 2 HAAR wavelets on a rotating surface.
===>>> What was the secret Wavelet?
Base level was to construct Pseudo-Wavelet based on the Sokolov CAM shape. Filter out the shape and then recognize the location of the "tape" - an induced fault.
===>> How did this fit into existing Wavelet Theory?
1. Yves Meyer produced a continuous Wavelet.
2. Ingrid Daubechies produced first recursive discontinuous Wavelet
3. Pseudo-Wavelet produce effective behaviour filter such as dynamics.
I was in 1996 inspired by Daubechies db4 wavelet and its db1 equivalent to Haar Wavelet. This allowed my construction of a "Shape based" (CAM) Wavelet.
==>What did CSIRO lose?
1. Lead in emergent Wavelet field.
2. Sokolov CAM system as applied to wider industry.
3. CSIRO lost the ownership of the critical part of the fibre processing machine.
===>> Read More https://lnkd.in/g3zVeiG4
It amazes me that people do not use suitable wavelet - HAAR ( or db1) to describe a periodic wave that is square. ( two Haar functions back to back ) or my own work in using "Pseudo Wavelets" originated from a "Shape based" transform. It would result in less Harmonics and no "ringing" with a benefit of knowing locations of start and finish of rising pulse such as Haar function. Just a note - there is no such thing as a "clean signal" in both strength or period - all signals in nature "suffer from random noise" be it temperal or volume. As for "trapezoidal" signal Δt - rise will alway approximate a Δt→0 But they will always be be Δt+ε.
This type of Fourier analysis falls apart if the signal is digital communication - there is no such problem in wavelet analysis. Wavelets can differentiate digital signal. And as 3D can differentiate a "shape" using a "pseudo Wavelet" transform from background noise of a rapidly moving radar signature such as a plane or a rocket. My work from (1995-2001) after CSIRO redundancy ( I have to say that as they claim everything just by being there employee at some stage of my life - for instance anything to do with IP and patents I generated at CSIRO )
I did my work in 1996 but I still see this Gibbs type approach in signal analysis. Fourier is best used as algebra tool box in Wavelets.
I liked the graphics... try reconstruction using Wavelets.
Some intro notes ( I need to do more );
===>>> What was the secret Wavelet?
Base level was to construct Pseudo-Wavelet based on the Sokolov CAM shape. Filter out the shape and then recognize the location of the "tape" - an induced fault.
===>> How did this fit into existing Wavelet Theory?
1. Yves Meyer produced a continuous Wavelet.
2. Ingrid Daubechies produced first recursive discontinuous Wavelet
3. Pseudo-Wavelet produce effective behaviour filter such as dynamics.
I was in 1996 inspired by Daubechies db4 wavelet and its db1 equivalent to Haar Wavelet. This allowed my construction of a "Shape based" (CAM) Wavelet.
==>What did CSIRO lose?
1. Lead in emergent Wavelet field.
2. Sokolov CAM system as applied to wider industry.
3. CSIRO lost the ownership of the critical part of the fibre processing machine.
===>> Read More https://lnkd.in/g3zVeiG4
Localized events, noise variation in periods, random effects - one off events can lead to erroneous conclusions (or Nobel prize). It has been a long practice to Window the sampling time. This leads to other effects on representing the frequency in its domain. Wavelets or "event" shaped transformations are the next step in analysis of data. Wavelets or Pseudo-Wavelets lead to ability to stretch the "window"or a filter to suit the scale of what we wish to focused on; "continuity" or "discontinuity" of the observed data. Fourier is still a great tool in proving that many Wavelets work. one can move Fourier from "simple"frequency analysis to "Fourier algebra" to support many many more previously invisible effects into the realm of observable. This applies to more than one dimension ie from a simple signal to a picture to a moving video to projections of other dimensions.
the most useful "simple Fourier" is reconstruction with minimum of constants - based on requirements of the application. In the 1990's I pointed out that some of this is like scientific "tea leaf reading". For instance one can recognized "square wave" from Fourier projection only if it is repetitive but not a digital signal - just an example.
6th =sixth (6th/8) of the "Source Books" CNC; G-CODE, C++ &MASTERCAM, I used in 1992 to manufacture a solve mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent
==>>How was this Useful???
In CSIRO talk and meetings are cheap.
The meetings can go on ad nauseam (everyone has ideas, but no skill set or eng science grounding).
I told the group leader that I solved the problem of high speed combing, being CSIRO- my mental health suffered dealing with his attitude.
==>>Did CSIRO Div of wool Tech had the manufacturing facility???
NO THEY DID NOT.
What they had was a CNC machine lacking software & operated from the front panel.
There were no USBx ports in those days only RS232/485 ports
The previous site engineer did not purchase the software needed.
Even if he did the memory within the CNC machine being DYNA DM4400 was limited allowing only a polygon to be made let alone a smooth CAM. I keep saying that its capacity was to make 50c coin replicas.
==>>What was CSIRO CNC DM4400 used for??
simple - the manufacturer in Taiwan made a G-code to manufacture a CSIRO logo. And that is what the machine did for many years before I repurposed it with software and hardware it.
==>> What was done to CNC DM4400 to extend it??
1. a user graphical interface CAD/CNC was written to write a G-Code
2. Large number of data was written IN/OUT taking an IBM-AT to load and unload CNC machine memory making it possible to produce smooth surface complementary Sokolov Cams.
3. In essence I written a MasterCam program and CNC interface.
==>> How did management behave?
I had little kids, I was working 12h a day weekends to make it happen. It was point of pride as my confidence from Barry Harrowield ( group leader ) was at all time low.
As result of the pressure I went on a sick leave. I was at home but I keep writing the software at night.
==>>So what happened when I came back from sick leave?
1.I implemented a successful CNC DM4400 upgrade.
2.We manufactured a first set of Sokolov Cams
3.The speed trials - now exceeding NSC Schlumberger specification repeated the experiments.
4.I was not given a promotion - year later I was redundant. ( lack of promotion, theft was complete )
5.Rumours started that I was on sick leave to manipulate redunduncy.
==>> How long did this take?
The combing project Dr.B.Kirby and I took 3 years from zero to a IP and patent. However CSIRO took 12y to lose it all with zero return using their "A-team".
===>> Read More https://lnkd.in/g3zVeiG4
Theft of recognition was completed.
The hurt was that they failed to deliver on what would have been beneficial to Australian Wool and Fiber growers and would have created a whole ne industries:
1. Manufacturing of the Australian Comb
2. Servicing of the Comb
3. World wide involvement in fibre processing
4. Surpassing The NSC Schlumberger in worsted combing.
IT CAN STILL HAPPEN
5th = fifth (5th/8) of the "Source Books" -FORTRAN & MSC ADAMS, I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
==>> Why was it useful?
Originally I used ALGOR Kinematic simulation but that had problems and wasted 3 months of getting a copy and trying to add CAM shapes or motion synthesis into the model. It took another 3 months to get software package needed and that was MSC ADAMS for full dynamic synthesis.
==>>Watch ADAMS sim Apron: https://lnkd.in/bdxvi57
==>>Watch ADAMS sim Front comb insertion https://lnkd.in/gZcUsBq5
MSC ADAMS was written in FORTRAN and needed that I relearn its syntax. But it had one advantage it could model "bounce" and obtain the limitation of speed for components used in comb. It also allowed me to reverse engineer a CAM or cams - you will notice that in Front comb insertion and feed forward we have 3 cams controlling motion. FORTRAN subroutine was added to data description of how to separate motions into 3 components which kept moving.
I obtained a FORTRAN from CSIRO budget but they were unwilling to pay for MSC ADAMS licence - so I got the copy from the university a copy not limited for the student edition. It was installed on my home computer to confirm software I was developing on the CSIRO site.
==>> How long did this take?
The combing project Dr.B.Kirby and I took 3 years from zero to a IP and patent. However CSIRO took 12y to lose it all with zero return using their "A-team".
===>>> What was next after this?
Generation of G-code for CNC production.
There was a problem in this CSIRO Div of Wool Tech had a CNC machine but no software to generate G-code. Like many things in CSIRO they claimed high tech but had no software - so I had to write it. I will cover that in the next post. DM4400 CNC could only make a CSIRO logo.
Behind the scenes I was working 12h a day as this was not the only part of my Combing project.
1. Create a numerical model for 1968 Belin & Verhagan theory
2. Create a numerical model for 1991 Kirby & Sokolov model
3. Create a proof that it was valid numerical model.
( I will cover that in another post, by this stage the software work was growing and I needed a way of linking it all together known as a program list or Worsted Comb Design package...will cover that in another post )
==>>This only took less the a year from scratch.
I was ready for the next phase - CNC manufacturing.
===>> Read More https://lnkd.in/g3zVeiG4
4th = fourth (4th/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
===>> This is 1991 Tech in CSIRO I used to take digitize the cams into CAD package. Prior to this we discussed the derivative analysis of the cams. Pointing out that cam size is unimportant as a result of first derivative size constant disappears. I needed to know exactly what NSC Schlumberger was doing during a worsted combing cycle on a dynamic and derivative basis. To accomplish this I needed to import and export data from a CAD package. For that we needed to learn LISP as it came with AutoCad 12 onward.
AutoLISP was derived from an early version of XLISP, which was created by David Betz.[4] The language was introduced in AutoCAD Version 2.18 in January 1986 until 1995 when it was neglected
===>>>All things in life are simple only fools overcomplicate.
They do this to cover the fact they do not understand the technology - and this technology was new in the 1991. there were no shortage of "fools" in management positions with PhD or hanger owns with only a MSc in unrelated area of science. This is a failure of CSIRO management of science and eventual loss of significant technology to NSC Schlumberger.
===>>Why was all this programing in LISP necessary?
1. accuracy
2. Ability to do a quick design in Polar coordinates using polylines.
3. Export into my C language Fourier analysis and filters.
I needed to import and export numbers on a unprecedented scale into and out of CAD in both polar and rectangular coordinates. My superiors "Physicist" had no idea of what I was doing other then "he is doing engineering things". The number exchange had to be accurate. I created several programs GraphXY.lsp, GraphRTheta.lsp...
===>>> What was next after this?
1. Analysis
2. generation of G-code for CNC production
Behind the scenes I was working 12h a day as this was not the only part of my Combing project.
1. Create a numerical model for 1968 Belin & Verhagan theory
2. Create a numerical model for 1991 Kirby & Sokolov model
3. Create a proof that it was valid numerical model.
( I will cover that in another post, by this stage the software work was growing and I needed a way of linking it all together known as a program list or Worsted Comb Design package...will cover that in another post )
==>>This only took less the a year from scratch.
I was ready for the next phase - simulation in ADAMS, CNC manufacturing.
AutoLISP was derived from an early version of XLISP, which was created by David Betz.[4] The language was introduced in AutoCAD Version 2.18 in January 1986, and continued to be enhanced in successive releases up to release 13 in February 1995. After that, its development was neglected by Autodesk in favor of more fashionable development environments like Visual Basic for Applications (VBA), .NET Framework, and ObjectARX. However, it has remained AutoCAD's main user customizing language.
Third (3rd/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
===>> This is 1991 Tech in CSIRO I used to digitize the cams.
Pre-requsit of reverse engineering of cams is ability to digitize the shape and extract the derivatives. I needed to know exactly what NSC Schlumberger was doing during a combing cycle.
====>>What is important?
1. the size of the cam was unimportant as Size of cams disappears in a first derivative. It was important to digitize the cams in such a way that no derivative is swamped with noise.
2. I needed to feed the Cam shape into the CAD package.
3. This was just Tech work and not Sokolov Cam System: https://lnkd.in/grqNXQ5
===>> How was it done?
1. C ( no C++ ) back then that I could access. I used
Ref: Numerical Recipes in C book set: Numerical Recipes in C: The Art of Scientific Computing
Publisher : Cambridge University Press; 2 edition
Language : English
Hardcover : 1020 pages
ISBN-10 : 0521431085
ISBN-13 : 978-0521431088
===>>> Copy https://lnkd.in/gha5PE8p
2. Data acquisition of shape: This was mainly as D/A cards had programs in "MS Assembler" or C.
3. Data needed to be "filtered" and derivatives produced - for that I used the Fourier method provided by this reference. It produced all of the derivatives I needed. Since in Fourier - differentiation is a simple multiplication.
4. I truncated the Fourier constants to reconstruct the shape of the cam, higher frequency orders were deemed as being unrealistic.
5. For purpose of reconstruction - or filtered shape was feed into a AutoCad 12/13 using LISP - this produced smooth shapes I used in ADAMS with FORTRAN subroutine assistance. Next- dealing with CAD and CNC reconstructions.
====>> What should be noted!
I written most of this software for Digitizing and built a specific Digitizing table amounting to about for this session 23 programs. Some in Pascal, mostly in C /C++, assembler, FORTRAN and LISP.( I will present LIPS & CAD as separate topic ). This also excludes my Icon driven simulation developed in 1991, to demonstrate dynamic difference between my design and NSC60 design. Key ingredient in stopping CSIRO to hand over my work to Schlumberger for free or ignorance as my group leader chose to believe NSC rather then CSIRO employee (myself as an expert)
====____>>>>This only took less the a year from scratch.
I was ready for the next phase - simulation in ADAMS, CNC manufacturing.
===>> Read More https://lnkd.in/g3zVeiG4
Third (3rd/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
===>> This is 1991 Tech in CSIRO I used to digitize the cams.
Pre-requsit of reverse engineering of cams is ability to digitize the shape and extract the derivatives. I needed to know exactly what NSC Schlumberger was doing during a combing cycle.
====>>What is important?
1. the size of the cam was unimportant as Size of cams disappears in a first derivative. It was important to digitize the cams in such a way that no derivative is swamped with noise.
2. I needed to feed the Cam shape into the CAD package.
3. This was just Tech work and not Sokolov Cam System: https://lnkd.in/grqNXQ5
===>> How was it done?
1. C ( no C++ ) back then that I could access. I used
Ref: Numerical Recipes in C book set: Numerical Recipes in C: The Art of Scientific Computing
Publisher : Cambridge University Press; 2 edition
领英推荐
Language : English
Hardcover : 1020 pages
ISBN-10 : 0521431085
ISBN-13 : 978-0521431088
===>>> Copy https://lnkd.in/gha5PE8p
2. Data acquisition of shape: This was mainly as D/A cards had programs in "MS Assembler" or C.
3. Data needed to be "filtered" and derivatives produced - for that I used the Fourier method provided by this reference. It produced all of the derivatives I needed. Since in Fourier - differentiation is a simple multiplication.
4. I truncated the Fourier constants to reconstruct the shape of the cam, higher frequency orders were deemed as being unrealistic.
5. For purpose of reconstruction - or filtered shape was feed into a AutoCad 12/13 using LISP - this produced smooth shapes I used in ADAMS with FORTRAN subroutine assistance. Next- dealing with CAD and CNC reconstructions.
====>> What should be noted!
I written most of this software for Digitizing and built a specific Digitizing table amounting to about for this session 23 programs. Some in Pascal, mostly in C /C++, assembler, FORTRAN and LISP.( I will present LIPS & CAD as separate topic ). This also excludes my Icon driven simulation developed in 1991, to demonstrate dynamic difference between my design and NSC60 design. Key ingredient in stopping CSIRO to hand over my work to Schlumberger for free or ignorance as my group leader chose to believe NSC rather then CSIRO employee (myself as an expert)
====____>>>>This only took less the a year from scratch.
I was ready for the next phase - simulation in ADAMS, CNC manufacturing.
===>> Read More https://lnkd.in/g3zVeiG4
===>>Addendum to previous post and origin of Australian Unique design known as "Sokolov Cams" and basis of the only ever Australian designed Worsted Comb a project I started at CSIRO. I could possibly add a whole chapter in "How to design Cams using Fourier techniques" and "Design Cams that solves "spring surge"at high speed by incorporating dampening elements in design of cams. Pay attention to the last design presented in this section "Modification and Uses for Basic Types of Cams" - this excludes a "German Cam Design" until 1991 considered to be most efficient cams. In simple terms Cam design has lagged from more modern mathematical approach that was used by myself.
For starters - although the "German Cams" involved cyclic behaviour to produce "True cycloid" many of processing requirements do not facilitate those goals.
Sokolov Cam System: https://lnkd.in/grqNXQ5
My work in 1991 brought full dynamic derivatives into Cams by designing and incorporating Fourier analysis & linear algebra as one would use in signal analysis. This approach greatly enhances the design aims of Cams.
Second (2nd/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
the pages are an extract from:
===>>> Looking for a source mechanism? (2nd/8 of my ref mechanical encyclopaedias, Yes I have a Library at CRD P/L exceeding CSIRO & some universities on my topics of interest. )
Ref: Illustrated Sourcebook of Mechanical Components
Publisher : McGraw Hill; 1st edition (30 April 2000)
Language : English
Hardcover : 1056 pages
ISBN-10 : 0070486174
ISBN-13 : 978-0070486171
Dimensions : 22.35 x 6.63 x 28.7 cm
===>>> Copy https://lnkd.in/gyFTqcNP
===>> Read More https://lnkd.in/g3zVeiG4
Second (2nd/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
I will go through each reference and describe the state of the knowledge at that time.
===>>How did I know "Sokolov Cam system" was original?
Reference and research of existing knowledge.
The best cam design in 1991 was known as "German Cams" - based on cycloid continuous derivatives. Much like Fourier addition - Sokolov Cams exceed "German cams" in design and dynamic response. I will discuss that in my next post by presenting 3rd reference book I used on that topic.
Sokolov Cam System: https://lnkd.in/grqNXQ5
===>>> Looking for a source mechanism? (2nd/8 of my ref mechanical encyclopaedias, Yes I have a Library at CRD P/L exceeding CSIRO & some universities on my topics of interest. )
Ref: Illustrated Sourcebook of Mechanical Components
Publisher : McGraw Hill; 1st edition (30 April 2000)
Language : English
Hardcover : 1056 pages
ISBN-10 : 0070486174
ISBN-13 : 978-0070486171
Dimensions : 22.35 x 6.63 x 28.7 cm
===>> New version
===>>> There are 3 aspects of cams and in particular "Sokolov Cams".
Prerequisite is Digitization of Cam derivatives using Fourier filters.
1. Calculation of desired derivatives Math synthesis.
2. Computational CAD shape that adds "cams" size to derivatives.
3. Manufacture CNC precision and assembly
This is followed by testing.
===>>> My work from zero- to patent was done in less than 3 years.
After that CSIRO wasted 12 years of trying to pick up where Dr.Kirby and I were made redundant for those with "better experience and qualification" to take our process forward - THEY UTTERLY FAILED and NSC Schlumberger was amused ( ok they laughed) at the technology that just fell into their lap.
CSIRO after stealing our recognition spouted in 1996 -technobabble words like "servo motors" & "computer control", but that was their fuel in how to secure funding - but none of it worked as they failed to understand the fundamentals of why Cam operation is superior in delivering combed worsted fibre. Yes technology is useful - but I can deliver on that NOT CSIRO selected people.
===>> SO what happened to incompetence in CSIRO
It was rewarded with CSIRO medal and fellowships. I also believe one of the protagonists became an adviser to the Science and Tech Minister.
==>> Lesson: incompetence is rewarded.
===>> Read More: https://lnkd.in/g3zVeiG4
==>>What to do with old Computers?
I never disposed of any of my computers past my CSIRO employment.
Most of my old computers follow an "upgrade strategy" for its MB until we reach a limit either by software (OS) or hardware (P4..i7 support), then we upgrade to a new MB based computer. That way on retiring the old computer - it is optimized and ready for storage and reference.
==>>> Why keep them?
There is no real need since they run as a image on my Virtual Computers. But occasionally there is an issue with Hardware simulation - its not perfect.
For example: Mass storage pre DVD/CD Burners.
In past I used a 250mb digital tapes to store everything post 1990. The earliest tape is labeled 1994. In 1996 I used them to backup my PhD research on Wavelets, Australian Comb Development and Sokolov Cam systems.
I am trying to close those specific loops in VM.
PS//I use some Laptops as Linux servers ( being ex-windows systems they can be re-imaged back into hardware and with SSD they are fast)
===...>> Where can I store copies?
I lived through the storage development technology and failures.
CD/DVD initially were made to be "forever tech" however few years in I discovered that they scratch and the coatings become less reliable. Also there were news that "fungus" and bacteria in tropics has started to attack the lining of the disks.
==>>What was needed?
A reliable redundant storage facility - that I owned. A solution presented itself in form of RAID5+ in our NAS server strategy. I have built a 16-64Tb storage facility.
It was proven reliable for the last 8 years and scalable as our storage needs grew.
===>> What is my AIM?
Most of my work and every visited R&D site in the 1990's to today have been "reaped". I need to go back to the backup tapes and load them into our NAS server. This also includes my website development prior to LinkedIn.
====>>> Why do I need this?
with rise of AI and private owned LLM one needs to give it a great knowledge base. AND THIS IS IT. My LLM has been given rules to "talk or intergate" GPT4 , but not to down load it into its own LLM model. It is way I will ensure a focus of our AI.
One (1st/8) of the "Source Books" I used in 1991 to solve a mechanical problem for CSIRO, a 200y old problem in Mechanical Engineering science. 60y old problem for CSIRO, NSC Schlumberger (France/Germany) & St.Andrea (Italy) that I solved resulting in a patent.
I will go through each reference and describe the state of the knowledge at that time.
===>>How did I know "Sokolov Cam system" was original?
Reference and research of existing knowledge.
The best cam design in 1991 was known as "German Cams" - based on cycloid continuous derivatives. Much like Fourier addition - Sokolov Cams exceed "German cams" in design and dynamic response. I will discuss that in my next post by presenting 2nd reference book I used on that topic.
Sokolov Cam System: https://lnkd.in/grqNXQ5
===>>> Looking for a source mechanism? ( 1/8 of my ref mechanical encyclopaedias, Yes I have a Library at CRD P/L exceeding CSIRO & some universities on my topics of interest. )
Ref: Mechanisms and Mechanical Devices Sourcebook Hardcover – 1 March 1991,
Product details
Publisher : McGraw-Hill Inc.,US (1 March 1991)
Language : English
Hardcover : 448 pages
ISBN-10 : 0070109184
ISBN-13 : 978-0070109186
===>> New version
===>>> Next Post will be on my other references on Cam mechanisms and design. It also deals with synthesis of dynamic derivatives.
===>>> There are 3 aspects of cams and in particular "Sokolov Cams".
1. Calculation of desired derivatives Math synthesis.
2. Computational CAD shape that adds "cams" size to derivatives.
3. Manufacture CNC precision and assembly
This is followed by testing.
===>>> My work from zero- to patent was done in less than 3 years.
After that CSIRO wasted 12 years of trying to pick up where Dr.Kirby and I were made redundant for those with "better experience and qualification" to take our process forward - THEY UTTERLY FAILED and NSC Schlumberger was amused ( ok they laughed) at the technology that just fell into their lap.
CSIRO after stealing our recognition spouted in 1996 -technobabble words like "servo motors" & "computer control", but that was their fuel in how to secure funding - but none of it worked as they failed to understand the fundamentals of why Cam operation is superior in delivering combed worsted fibre. Yes technology is useful - but I can deliver on that NOT CSIRO selected people.
===>> SO what happened to incompetence in CSIRO
It was rewarded with CSIRO medal and fellowships. I also believe one of the protagonists became an adviser to the Science and Tech Minister.
==>> Lesson: incompetence is rewarded.
===>>> Why so much interest in worsted combing?
many people have stated to me including my wife:
"why do I have such unhealthy obsession with worsted comb?"
"you have so many other skills and expertise in CFD, FEA, 3DP, CNC, Wavelet Math and programming both in hardware and software...not to mentioned remote operated drones... your lab exceeds some of the university and csiro ...why bother?"
==>>To answer those questions...
1. Worsted Comb is most challenging machine I have ever encountered.
2. I love dynamics and there is nothing more humbling when you design and use of exceptional level of math -and it turns out not to work
3. Reverse synthesis of mechanical motion is challenging due to number of derivatives it contains.
4. everyone has an opinion at how it could be done -but do not have the know how to implement or understand the solution.
5. I like to finish what I started - especially if I have a solution
6. Worsted Combing is a project of National Interest to Australia
7. it has something to do with my skills &at cycle frequency of 5Hz level
Example jpg: this is a front comb motion - and its possible variation between PB29-1972 and PB31L-1992. The NSC Schlumberger engineered the synthesises of motion but not the dynamic behaviour of the motion in 1991 I overtaken their efforts in cam design, by adding dynamic synthesis producing cam shapes known as "Sokolov Cam" technology. What tickles me pink is fact that in Australia I have overtaken Schlumberger (France-Germany) &St.Andrea (Italy) in Machine design science. Statements by CSIRO management "how difficult is it to alter a cam shape" betrays their ignorance of the skills needed.
Secondary to all of the above is the insult to intelligence of working scientist and engineers that CSIRO management of science commits.
A lot of projects in CSIRO are not exciting they do not contain current buzzwords or shiney new technology BUT are projects like Worsted combing. Because of buzzwords, shiney stuff science and technology, tech that is in national interest suffers. In pre 1995 we in CSIRO got paid to look at new tech as part of our job, not job definition.
The scientist that realized technical skills incompetence move into "management of science" in many places like CSIRO. They keep the paperwork moving while muting their failures. This leads to their ; "never mention that project again" mentality.
==>> Technology beneficial to Australia is ignored
This technology would not be ignored if I fulfilled some political niche or patronage. that is how "management of science" works in CSIRO.
===>>>Recommended
===... Download>>> https://lnkd.in/gXYCDmWb
Make your own AI to aid in your program objectives.
I have a large database stretching 35y of research occupying 14Tb, including all the correspondence, interactions AND help files I created for CAD, 3DP , CNC solution to worsted combing theoretical fibre performance and manufacturing in existing NSC Schlumberger Comb and my work at CSIRO ( which is now less of an IP but more a historical reference ) .
I initially started collecting my data based around a need to upgrade my computers at regular intervals keeping my work centralized on a RAIDx NAS server. With 10Gb and above communication - the server feels more like a HDD in any of my 12 computers. Making R&D computing feasible in real time. It is only natural to implement an AI to manage "all" of my R&D needs.
Those are just necessary tools of the 21st century - what in not necessary is parasitic management positions.
Yes, my computer network does play a mean chess game. I am looking to update its chess board in which it can move pieces directly - much like Ghost chess.
====>>>>Lesson that CSIRO needed to learn is this; IF YOU HAVE A SUCCESS DO NOT CHANGE THE PEOPLE THAT GOT YOU THERE.
I came from a disadvantaged background and was never destined to occupy "management positions". I remember working in an meat processing plant and cleaning the boning tables with steam and mucking out the hide tanning yard while watching the "supervisors" sitting in a air conditioned site office playing cards - I wanted to be there. Did not realized that one needed an excellent communication and writing skills, not provided by Australian Technical school system- everything else was optional and can be fudged and spun into a narrative. This all changes with introduction of AI.
It has taken me 15y to build my facility using many times 2nd hand equipment discarded from places like CSIRO, as they were usually chasing that "shiny new" opportunity for media attention - for which results were irrelevant and inconvenient. They just needed to get into the "word spin", here it is deployment of LLaMA ; AI deployment. My AI will be biased to my opinion and goals.
GPT-Engineer is insane ?? It's an AI agent that can write an entire codebase with a prompt and learn how you want your code to look:
? Asks clarifying questions
? Generates technical spec
? Writes all necessary code
? Easy to add your own reasoning steps, modify, and experiment
? Lets you finish a coding project in minutes.
↓
Are you technical? Check out https://AlphaSignal.ai to get a weekly summary of the latest research and breakthroughs in AI. Read by 100,000+ engineers and researchers.
This sums up my breakthrough in 1991, leading to producing a more efficient Australian Worsted comb. This was an utter failure of CSIRO to capitalize on the R&D advantage.
How stupid is management of science in CSIRO?
==>>Opportunities wasted...with no regrets
==>> CSIRO notables then write books on innovation
Conjugate Cam Tech as noted in "Cam Design &Manufacturing Handbook" by Robert L. Norton also one of my references. And recent "Cam Design Handbook" by Harold A. Rothbart (covered in previous post and publication). None of those handbooks carries a design of my solution to NSC Schlumberger (France) and St.Andrea (Italy) problem of increasing worsted comb productivity, known as "Sokolov Cams" and shown in a jpg
==>>>This is how simple good industrial innovation was in 1991
On the Left is the traditional Conjugate cam with a single pivot point. On the Right is "Sokolov Cam". Not described in any current textbooks on Cams and Cam manufacturing. It was done in 1991.
CSIRO decided to take out a patent on "Apron Retraction" and not the mechanism. The Sokolov cams - latter they gave the mechanism away free feeling smug that they solved retraction problem through my work but failed to observe that NSC Schlumberger has implemented Sokolov Cams on other Worsted comb mechanisms.
The Icon driven computer program was written by me in 1991 to assure textile equivalence in competing cam tech, Sokolov cams were superior and although I have altered the retraction I did much more - I increased the efficiency by being able to run the comb 40% above manufactures specification. Return to CSIRO was zero.
The other problem they failed in they did not recognized the power of software that was evolving in PC. The icon driven software was written by me using a Mathconnex software, much like Simulink in Matlab (in '91 it was all new stuff). Each icon represented a component and setting in a comb -this was retraction mechanism. I did this because Schlumberger convinced Barry that they can do what I did for the patent,. barry being a PhD in Crystallography fell for their ploy, what a du*b ar*s -he then tried to cover his mistake with my redundancy and that of my supporter Dr.B.Kirby. CSIRO Div of Wool then promptly went to sleep on a job wasting $millions, looking pompous and self assured; until they were disbanded for non-productivity. Not to worry they gave each other CSIRO medals
==>> Technology beneficial to Australia is ignored
This technology would not be ignored if I fulfilled some political niche or patronage. that is how "management of science" works in CSIRO.
Read More: https://lnkd.in/g3zVeiG4
Sokolov Cam System: https://lnkd.in/grqNXQ5
==>> Technology beneficial to Australia is ignored by CSIRO.
This technology would not be ignored if I fulfilled some political niche or patronage. that is how "management of science" works in CSIRO.
Read More: https://lnkd.in/g3zVeiG4
Sokolov Cam System: https://lnkd.in/grqNXQ5
Addendum I; 8th=eighth (8th/8) of the "Source Books" in EE I will do another Addendum for ME. THIS IS ONLY PRACTICAL STUFF!!
Lot of practical ME and Math- signal analysis I already covered.
There are many undergrad textbooks but those are practical reference books I update &own.
When I was a young man in the 1970's ,I went to a Technical school -my future was -"get an apprenticeship". I was grateful to have been streamed and my education was enriched. However that did not last long, during my HSC/TOP I encountered English as a subject. Once again excellent marks in : Chemistry, Physics and Math (Pure &Applied ) but did not do so well in English (also applies to many English native Tech graduates)
===>>>How I selected my first Undergraduate Degree?
I was one of the few people that passed Chemistry in 1976 HSC/TOY and scraped through in English, I was lucky to pass. My choice was Mechanical Engineering as it contained back then: Theoretical Mechanics, Thermodynamics, Math and some Electrical engineering. It also offered a clear path to employment as a Mechanical Engineer.
===>>> What was my choice of a second Degree?
It was a Electrical engineering -expand on my first degree.
===>> Why higher degree?
After the problems I experienced in commercializing a product we made in 1980's - scallop opening robot - CSIRO got interested in my skill set after applying for a job I was accepted. They were impressed.
However after a while working there I was just another guy with a degree and no PhD ( in Physics) and above all I was an engineer in a land of Physicist. Wow - do you not constantly get put down by people that a less creative and until I got there had very little success in IP and patents. BUT that did not stop them putting me down. To get a higher degree and be equal to a Physicist - it was no great effort. However I missed out on "Physics politics in R&D 101" so while I was producing results they conspired theft. My lack of communication and written skills played right into their hands - typically - "look he can't even write".
Their song was : "we have enough scribes - we need commercial results". Pity his words did not match his behaviour and actions. When I achieved my Ip and patent on combing - I became just another engineer producing results under a guidance of a "Great Physicist". My PhD is Math based Wavelets -Mathematicians are the only discipline Physicist steal from and respect. Chemistry people are despised by Physicist.
==>> Read More https://lnkd.in/g3zVeiG4
There is also several practical references to electric motors of all sorts from direct ac to hybrid... and their control circuits.
==>> Why MechEng?
One of the few original Engineering disciplines. Mechanical Engineering spun out many other disciplines - one of the earliest was EE. Now days there are many many more even a shortcut called; Mechatronics, a combination of ME+EE all in a 4 year course. I enjoyed my ME degree in late 1970 recognized in early 1980 that I needed an EE ( both Power and Electronics ). Many of this learning I did in my last 40y. I never regretted doing ME as it gave me a thorough grounding in Dynamics, Theoretical mechanics and Thermodynamics. It also gave me an ability to synthesize motion I desired.
==>>Why EE?
That was a natural progression from ME. More and more equipment came which controllers starting with PLC in the 1980 to more sophisticated microcontroller and full on computer control. There is a limit what a computer control can do at current of 20mA and voltages of 3.3, 5 or 12V. To control real machines one needs Power in shape of Current for Torque and Voltage for Energy potential. We all know that is some from of Power is proportional to Voltage multiplied by Current.
===>> Why Math?
Everything is Math and signal conditioning hence my work in Wavelets and Signal conditioning. There is no point in releasing power efficiently if there is no good reason to do so - and it needs to be proportional. Linear algebra and computing makes tricky problems - easy. FromWavelets, Fourier to Laplace - it changes differential equations in something we can solve and implement in practical machines.
===>>Why get a Job in CSIRO?
Many people that I learned from worked in CSIRO of the past. It was impressive. But by the time I got there and by the time I left they morphed into a "management of science" and become 1/4 of their size by 1995 the year of my redundancy.
===>> Why CSIRO is not an option for 21st century?
I wanted and won an ongoing position in CSIRO ( through IP and Patents ) because in the 1985 when I started - they had the tools and vision; computers and CNC manufacturing. They had money- rich kids with shinney toys. With CNC, 3DP &Computing becoming cheaper -you do not need to be their employee. The stuff -such as technical expertise and equipment is now in very short supply in CSIRO). They are there to manage media and comment on "everyone else's technology &new science". Life is short and once you get your degree or even a PhD. Go out there and self fund - monetary gains are more than a empty promises of the CSIRO.
My current Lab making the only ever Australian high tech fibre worsted comb.
====>>>> It's good to have variety of technical skills. This is what real ex-CSIRO scientist engineer does; most of us old hands that lived through the 1985-1995 (&beyond) have seen computers evolve from UNIX servers to Linux everywhere ( BNC to WIFI ). In process we developed real skills - like repairing and troubleshooting the equipment. ===>>>example: NAS server failure - replacement cost of new QNAP $3.2k ( no NAS 8gb HDD drives $250-300 each additional $2k ) and worry that the NAS RAID10 will not be compatible is nil. Repair cost little bit over $80. ===>>>What is most important? Loss of data on making, designing and simulating first ever Australian comb. This is a critical part of wool/textile fibre processing, my work with Dr.B.Kirby developed and had first patents in this area over NSC Schlumberger AND we still do. I live in hope that CSIRO will purchase my company for the IP to build in Australia a world's most advanced worsted comb. Other then that if I did this privately CSIRO would open a court case against my company for the intellectual property and technology ownership. Tragically they had a go with their "A-team"after our redundancy and failed - NOT all physicist and engineers are the same; that is all the found out. At least I would like a letter of release so that Australia can benefit from our technology. ====>>> HINT ..sorting box and a battery powered screwdriver is handy... I did most of this from memory BUT normally I take notes and pictures. I also did not look at the YouTube for "How to..." I did this by feel - after the item arrived I did look at the "YouTube" - but guess what - they did not have a video I needed and one close to my unit was different configuration. so there ...not everything is on the google and youTube - it come a time to use real skills. ===>> Read More?https://lnkd.in/g3zVeiG4 ==>>>Lesson that CSIRO needed to learn is this; IF YOU HAVE A SUCCESS DO NOT CHANGE THE PEOPLE THAT GOT YOU THERE There used to be people with real technical skills in CSIRO now they are gone replaced by - "management of science" but limited skill set. Tech personal that I learned from are also gone or dead like Phil Greig. #scienceandtechnology?#australia?#career?#TeamCSIRO?#science?#phd?#stemcareers?#stemeducation?#stem?#engineering?#csiro?#xcsiro =>>I once had respect for CSIRO now I don't?#engineersaustralia
==>> Read More https://lnkd.in/g3zVeiG4
Forever young!
I will be coming up to a 50y association with TKD - always room to improve.
alway people to lookup to.
Sunday morning dawn training...clears the mind and focuses the energy.
32 separate forms to follow + stretching + stick & staff practice
Great for mental and body health.
Technique above brute strength.
Respect to everyone.
One needs to understand time and meditation...
However in 1998 I was run over by a Toyota 4 wheel drive Landcruiser while on my motorcycle heading into Melbourne to teach - I never fully recovered from that encounter BUT I deal with its consequences. Learning to walk was a consequence...BUT I never gave up.