Document D5 for the Formal Enquiry
The "background notes" added in blue font were for the benefit of the Formal Enquiry Panel.
They were NOT part of the original 2005 warning letter.
Background note
In March 2005, Bill Courtney received a tip-off from within the University, that Dr Oyadiji et al had breached the confidentiality clauses of the PedSALi Collaboration Agreement, by publishing papers at two international conference ins the USA, without informing CI or Dow.
In early July 2005, Bill received hard copies of four papers from the University of Manchester. The wording of their titles gave no clues concerning when or where publication was planned. The cover note attached to the papers simply stated,
"Please find enclosed papers which have been prepared for publication from the PedSALi project."
The following letter was sent in reply.
CHESHIRE INNOVATION®
Engineering Consultancy
17 Vale Road, Timperley, Altrincham, Cheshire, WA15 7TQ
Tel/Fax 0161 980 5191, E-mail #######
Web site www.cheshire-innovation.com
11 July 2005
Dr. Olutunde Oyadiji, Senior Lecturer in Engineering
School of Mechanical, Aerospace and Civil Engineering
The University of Manchester
PO Box 88
Sackville Street
Manchester
M60 1 QD
Background note FOR WEB PAGE READERS
This letter was sent Royal Mail Recorded Delivery.
Email copies were also sent to
(i) six relevant personnel at the University.
(ii) The EPSRC link person for the PedSALi project.
(ii) The relevant Department for Transport and Foresight link people.
(All the University, EPSRC and other personnel are named on page 10 below.)
No reply has been received.
Dear Dr. Oyadiji,
Thank you for your letter dated 28 June 2005 and copies of four PedSALi papers the University has prepared for publication. Your letter was received, Special Delivery, 6th July 2005.
- Unfortunately, you omitted all of the essential details relating to where and when the University hopes to publish the papers. Please supply these.
- I have, with the aid of a reading assistant, carried out a preliminary review of the papers and regret to inform you, that all four are fundamentally flawed. They describe experiments that are weak on design and fail to present valid evidence of how Shock Absorbing Liquid (SALi) , performs under impact, as described in the published patent literature. Please see Appendix One for my initial analysis.
- The University papers cite the relevant patent literature, so it is surprising that these design flaws occurred.
- I am further surprised that the University has considered publishing such flawed papers because on numerous occasions, the University research workers were alerted to these deficiencies.
Sample references of these warnings are given in Appendix One.
- In a letter to the EPSRC dated 22nd March 2005, I explained why the commercial future of SALi Technology would be placed in jeopardy, if flawed University research results were published. Therefore, under the confidentially clauses 5.10 and 5.11 of the PedSALi collaboration agreement, signed by the partners , 16th March 2001, I request that these papers are not submitted for publication.
Owing to the serious nature of the research defects and the potential damage their publication could do to the commercial development of SALi Technology, the University must not ignore this request.
For your convenience, the Confidentiality clauses 5.10 and 5.11 are copied below as Appendix Three. Also, please find attached a copy of my letter to the EPSRC dated 22nd March 2005. - My University colleagues will be aware that I have suffered ill health due to the stresses of dealing with the University on the PedSALi and CrashSALi projects.
[Sample evidence that the University was aware of my health problems: Minutes, PedSALi Formal Meeting 9, 27 June 2003, Section 3.]
The continuing threat of litigation by the University, via their solicitors Eversheds has only exacerbated these stresses.
I am now suffering macular degeneration and have great difficulty in reading. This problem is currently being treated by the Royal Manchester Eye Hospital, but an improvement in the near future is not anticipated. If the University researchers wish to engage in a prolonged debate on the quality of their PedSALi papers, I would be grateful if you would send me copies of all four papers enlarged to A3 page size.
- The University researchers appear to conveniently change the supposed properties of Dow Sconapor e.s.p. beads when using them for Smart CrashSALi or PedSALi work.
In the CrashSALi work, the researchers used Sconapor e.s.p. as an elastic material, when testing a SALi based suspension system. One of the reasons the CrashSALi report was returned to the University Vice Chancellor, was that the University research workers had used the wrong materials. E.s.p. is not elastic, so it is unsuitable for use in a prototype vehicle suspension system.
Appendix Two is a recap of the relevant technical arguments made in my correspondence with the Vice-Chancellor.
It is interesting to note that in the second paragraph, page two of the PedSALi paper titled "Characteristics of a Deformable Beam Filled with a Shock Absorbing Composite" the University workers write, "On impact, the beads are plastically deformed. Therefore new beads were used for every test performed."
- How does the University justify ignoring my MP's plea for a suspension in its litigious action and continuing to press Cheshire Innovation, via the University solicitors, for full payment, for the Smart CrashSALi work, when its own research workers admit that e.s.p. beads are only effective for one impact?
- An Internet search suggests that the University researchers have presented papers on the PedSALi project in October 2004 http://www.saviac.org/75th_Symposium/Abstracts%20Final.pdf.
and March 2005. http://www.spie.org/Conferences/Programs/05/ss/conferences/index.cfm?fuseaction=5760
Were PedSALi papers published at these conferences? If so, why was this kept secret from Cheshire Innovation?.
Please provide a prompt and detailed response to the points made in this letter.
Yours sincerely,
- A. Courtney
c.c. Graham Brady MP, House of Commons, London, SW1A 0AA
Background note
In 2003 I was working on the PedSALi project as a Research Fellow at the Victoria University of Manchester (VUM).
At the same time I was a guest lecturer on SALi Technology at the nearby University of Manchester Institute of Science and Technology (UMIST).
In 2004, the two universities amalgamated to form a new university, The University of Manchester.
In general, there were excellent engineers at both institutions, but when it came to working on SALi Technology, the UMIST staff were more experienced and better equipped. They possessed more adaptable drop test rigs, a polycarbonate rod Split Hopkinson Pressure Bar and two Instron machines.
VUM had one Instron machine that had been broken for several years.
The UMIST engineers also possessed more appropriate skulls. They were crash experts, the VUM engineers specialised in vibrations.
It is understandable that the VUM staff would have felt vulnerable in the countdown to amalgamation.
A desire to cover up relative deficiencies may go some way towards explaining the nature of the failure of the PedSALi project.
But this is not an acceptable excuse for unprofessional behaviour.
APPENDIX 1. Key deficiencies in the PedSALi project research papers
1.1 Key deficiency in papers relating to determination of SALi core characteristics
Introduction
The aim of this project was to design and test a prototype soft, pedestrian friendly car bumper, to meet pending EU regulations, as they stood when Dow Chemicals approached Cheshire Innovation in 1999.
Figure 1 above illustrates the type of impact test that would be carried out by a vehicle manufacturer, towards the end of the project. Typically, this type of crash test costs about £20,000.
In order to minimise total crash test costs, the real test would have been preceded by computer simulations carried out by Dow. One of the contributions of the University researchers was to provide Dow with "core characterisation" data. Dow would have plugged this data into their software programmes, used to carry out the crash test simulation work.
The commercial challenge was to design a slim car bumper, with a curved front, that was soft for pedestrian lower leg impacts, but stiff for low speed parking type impacts. Dow considered that a material that met these apparently conflicting requirements would be a market winner. Existing foam and honeycomb materials were either soft or stiff and could not solve the conflict of stiffness problem.
Bill Courtney's MPhil research experiments suggested that a SALi filled bumper could meet the challenge, but his early experiments were restricted to small scale, low impact energies. They needed to be repeated, in far higher impact energy experiments before Dow could approach a vehicle manufacturer, with a view bringing them on-board, as a PedSALi research partner.
The unique properties of SALi were predicted to be more effective than foam filled bumpers during the early stages of pedestrian impact, but, if correctly packaged, stiff for parking type impacts. To get the design right, accurate core characteristic data was required on the early stages in the compression of the bumper, when it was only slightly compressed. This was the key to an effective, but slim bumper design.
Figure 2 above depicts the type of test cell used by the University, for the core characteristic work.
Defects in the University Core Characterisation Research
The 6.6 kg weight of the piston pre-loaded the SALi material, so that it was impossible to collect data that would help Dow simulate what would happen during the early stages of an impact.
A second problem was that a system consisting of a sample of SALi, covered by a heavy piston, buffered by a 50 mm thick slab of rubber, behaves very differently to a sample of SALi covered by a plastic bumper shell.
The best that can be said for data collected using this type of test cell is that it could be used for designing a bumper system comprising a normal bumper, with a 6.6 kg block of metal and a 50 mm thick block of rubber mounted on the front. This is not a practical proposition.
Background note relating to the experiment discussed below.
The aim of this experiment was to determine the visco-elastic properties of a composite material at different impact velocities and forces.
The ex-VUM University researchers believed that they could overcome the errors introduced by the massive piston and rubber buffer shown in Figure 2, by adding acceleration measurement devices to the apparatus, wiring the accelerometers up to a computer and programming the computer to do some clever algebra. This may be true for a pure elastic material, but not for one that has been specifically designed to exhibits visco-elastic properties. The authors have masked this failure in logic by presenting similar looking results for two different impact velocities, but with the viscous contribution being relegated to a minor role, by using a matrix fluid having a low viscosity, similar to water.
In contrast, UMIST engineers offered a different and valid test cell design. This consisted of a hollow Perspex cylinder coupled to a horizontal polycarbonate Split Hopkinson Pressure Bar by low friction O ring seals. The design was validated in quasi-static compression tests using UMIST’s Instron machine, shortly before the merger. The UMIST approach eliminated the loading error due to the gravitational pull on the piston, minimised the inertial mass error during impact and the coupling errors caused by having to use a rubber buffer.
It can be seen from document AC2, pages 9 -12, the VUM engineers successfully employed heavy handed and intimidating tactics to prevent my funding valid core characterisation research at UMIST. To appreciate the wider consequences of these tactics, it will be necessary to read the whole of document AC2.
A specific example from the University paper titled, "Core Characteristics of a Shock Absorbing Composite"
The University researchers are over-optimistic when they write, "At the same time as taking into account the piston effects, the effects of the rubber buffer are also eliminated." (Page 4, after equation 3.)
The laws of Physics that dictate the nature of the impact cannot be defied by monitoring the impact using accelerometers and solving a pair of simultaneous equations, to eliminate the elastic constant of the rubber.
The effects of the rubber buffer have not been eliminated by clever algebra. The algebra simply eliminates the requirement for the stiffness and damping of the rubber buffer to be explicitly known. This information is picked up implicitly by the accelerometers.
Why CI objects to publication: The innocent reader, attempting to use this "core characteristic" data to devise a mathematical model of how SALi filled packages performed under impact conditions, would be led completely astray.
[Sample evidence of CI's warnings on the flawed nature of this work: Minutes, PedSALi Formal Meeting 11, 30 January 2004, Appendix A]
1.2 Key deficiency in papers relating to impacts on SALi filled beams
Background note
The description of the PedSALi project on the DfT web site includes the following statement, "The packaging provides a flexible leak-proof envelope to contain the liquid and capsules, but, importantly, it must not stretch significantly during load application."
(http://www.rmd.dft.gov.uk/project.asp?intProjectID=10434 Paragraph eight)
The University researchers ignored the essential packaging requirement and carried out a series of experiments in which the package was intentionally allowed to stretch during impact. It therefore came as no surprise when their research findings underestimated the effectiveness of the SALi mixture in absorbing impact energy.
The University researchers have presented results relating to impact tests in which an elastic/plastic packaging that expands under impact is filled with a plastic composite material that compresses under impact. In certain experiments, constraints were used, to prevent the packaging stretching and bulging, away from the impact zone, but no attempt was made, to reduce the stretching of the packaging material, in the close proximity of the impact zone.
The manner in which (i) the packaging expanded and (ii) the filling compressed, under impact conditions were unknown. This means that the researchers embarked on a series of experiments, which violated the principles of good experimental design, by needlessly introducing two unknown variables, of comparative impact energy absorbing importance.
In the soft sciences, such as group psychology, where multiple unknown variables are unavoidable, the analytical problems are reduced, but not eliminated by using statistical instruments. There is no evidence that the researchers refined their results using statistical instruments.
Cheshire Innovation (CI) offered to work with the University researchers, to design a packaging that would overcome the two unknown variables problem. CI supplied low stretch sheaths for minimising the problem. There is no evidence from the papers that the low stretch sheaths were used in the experiments. The University research workers did not report back to CI, on any possible problems relating to the use of the sheaths.
CI's published patent literature lays heavy emphasis the importance of using stout, flexible packing that does not stretch significantly during impact. This literature also emphasis the importance of the SALi packaging having the optimum vertical geometrical shape, at right angles to the length of the car bumper.
The traditional D shaped car bumper cross section, as depicted in Figures 1 and 3 approximates to the optimum design.
In contrast, SALi filled beams, as tested by the University researchers, have a circular cross section. This is a weak design shape that undersells in impact energy absorbing potential of SALi filled beams. CI constructed a former that could be used to restrain the beams in a D shape. There is no evidence from the prepared papers, that the former was used in any of the University experiments.
About the best that can be said for the University research results, is that they may be useful in designing a car bumper, having an unconventional circular cross section, with part of the bumper facia being removed in the region of the lower leg impact test.
Why CI objects to publication: The SALi filled beam experiments are bad science. Their publication would damage the good name of SALi Technology.
[Sample evidence of CI's warnings on the flawed nature of this work: Minutes, PedSALi Formal Meeting 10, 29 September, 2003, Section 6]
1.3 The costs of the failure of the PedSALi Project
The so called soft bumper, as originally planned, as an EU legislative requirement, would still inflict painful injuries on a pedestrian in a 40 km/h accident.
The watered down EU requirements for a pedestrian friendly bumper will result in more crippling injuries, compared with the earlier proposed legislation.
The Cost: Due to the failure of the University to deliver meaningful results in a timely manner, Dow and Cheshire Innovation have lost a market opportunity. But the real losers may be tomorrow's pedestrians.
Background note
The EU watered down its soft bumper requirements, but did not completely abandon then. It proposed that soft bumpers as originally demanded would be required for all cars manufactured from 2012 onwards. But only if the conflict of stiffness problem could be solved.
By publishing false data that implied that SALi filled bumpers could not solve this problem, the researchers were denying a future generation of pedestrians the chance of soft SALi filled bumper protection in order to hide their own bad research.
Appendix 2
The DTI Smart supported CrashSALi project:
Introduction
Prior to commencement of the CrashSALi University research work, Bill Courtney and the University principal investigator visited the research headquarters of the Malaysian Rubber Producers Research Association (MRPRA) near Luton., 5 March 2001, to seek their advice on the best materials for multiple impact formulations of SALi. MRPRA made two recommendations that were included in written plan of work submitted to the University, at the beginning of the project.
These were:
- The elastic foam capsules for use inside the unit must have full and rapid recovery properties. These properties must be maintained over several million cycles. Natural or synthetic closed cell foam rubber was recommended. A suitable block of rubber, for dicing up into capsules was supplied by MRPRA.
- The matrix fluid must be chemically compatible with the rubber. Hydraulic fluid, commonly available as brake or clutch fluid was recommended. Lubricating oils should not be used, because they react corrosively with rubber.
The University researchers chose to ignore the plan of work supplied by Cheshire Innovation. Instead of foam rubber, they used expanded polystyrene bead capsules. Expanded polystyrene is adequate for use in car bumpers, that are only required to take one impact before the bumper is replaced. But, because of its brittle/plastic nature and slow recovery properties, it is unsatisfactory for use in vehicle suspension systems.
Instead of hydraulic fluid, the researchers used corrosive engine oil as the matrix fluid. Prior to handing the rubber bellows, for use in the prototype unit, to the researchers, the rim was spot tested, to check on MRPRA's predictions. The tests confirmed that brake fluid was compatible, but that engine oil was corrosive. A written statement and photographs of the corrosion experiments were handed over to the researchers, with the bellows.
Appendix Three
PedSALi Collaboration Agreement signed by authorised representatives of The Dow Chemical Company Limited (DCCL), The Victoria University of Manchester and Cheshire Innovation, 16th March 2001
The relevant agreement clauses:
5.10 The obligations of confidentiality and non-use set out in this Agreement shall subsist for the duration of this Agreement and for a further period of five (5) years.
5.11 In the event that the University wishes to make any publication regarding the results of the Project the University shall submit such publication to DCCL and Cheshire Innovation not less than thirty (30) days prior to the intended date of such publication and in order that to (sic) DCCL and/or Cheshire Innovation may request that such publication be amended in order to protect commercially sensitive information proprietary to DCCL and Cheshire Innovation. The University should receive notification of any objection to the proposed publication prior to the intended date of such publication. In the event that the University does not receive such notification it shall be entitled to proceed with such publication without prior consent.
Details of email letter sent on same day as the hard copy.
From: Bill Courtney <###########>
Date: 11 July 2005 20:344
To: Tunde Oyadiji #######>
Cc: Richard Bailey @ epsrc.ac.uk <######## @epsrc.ac.uk>;
Cate, Peter (PJ) <###dow.com>; CHRISTINE ASPEY <####@nwda.co.uk>;; Dr John T Turner <#####.man.ac.uk>; Dr. Zhu <#####>; Eugenio Toccalino <#####dow.com>; GEORGE GEORGIADES <#####.man.ac.uk>; john.j.harrigan #### Michelle Cooper <#####maninv.com>; Nick Barter <####>; Pat Selwood <####>; Prof Jan Wright <#####.man.ac.uk>; Prof. Steve Reid <######>
Subject Proposed University of Manchester PedSALi research papers.
Dear Dr. Oyadiji,
Please find attached an electronic copy of a response to your letter dated 28 June 05. A hard copy is in the post.
Also attached is a related letter to the EPSRC.
Yours sincerely,
Bill Courtney