Agenda of the organizing session of the group charged with validating certification corridors developed for the new Flex-GTR lower legform test tool to be introduced in Phase 2 amendments to GTR 9.

Agenda was agreed without any change.

Proposed scope, objectives, rules and procedures for the Task Force for the Review and Update of the Certification Corridors (TF-RUCC) developed for the new Flex-GTR lower legform test tool to be introducted through phase 2 amendments to GTR 9 on Pedestrian Safety.

The intention of this meeting is to provide an understanding to TF members of the calibration procedures. Technical discussions will start in January.

This TF group is dedicated to certification corridors only. If different technical problems occur, IG GTR9-PH2 decides whether new TFs will be formed or not.

Chairperson (Atsuhiro Konosu) presented terms of reference for TF-RUCC-K-03. All GTR certification procedures are to be reviewed and to be updated if needed.

Bernard Been: In the current schedule there is no activity of Round Robin testing. We do not want to make the same mistake (rush to make certification corridor with limited data) and must carry out Round Robin, test 10 legs with current material batches from at least 5 labs. Oskar Ries supported this statement.

It was asked when 10 legs could be made available. This would depend on customer co operation when they can they spare legs for testing.

Bernard Been: it is possible to complete round robin in 21 weeks if well managed allowing for transportation and efficient testing, e.g 4 legs at a time.

Carsten Hohmann: how many legs have been made? Mark Burleigh answered 23 customer legs have been built, and 5 prototypes (3 JARI, 2 Humanetics).

Round robin legs should be certified (component and assembly level) at Humanetics before Round Robin testing to ensure consistent spec is present in all legs.

Oliver Zander: can we look at old data and not carry out expensive testing?

Steve Pingston: do we have agreement to shift corridors?

Chairperson: technical details to be discussed later, component level corridors need to be reviewed first.

What are we to achieve by end of March finalization or draft corridors?

Steve Pingston: what is the goal of this task force?

Dr Konosu: The goal of the RUCC is to make desirable corridors and go step by step with confident data, first focus on component level (tibia, femur, knee), then full assembly by end of March if possible.

Agenda for the first regular session of the RUCC Task Force (following the kickoff session held in November 2011).

Chairperson stated objectives of the 1st TF-RUCC meeting as follows:

• ► Review current situation.
• ► Decide how to make progress of this TF-RUCC activity
• ► Develop working plan with time schedule.

TF members agreed on that.

Chairperson stated objectives of the 1st TF-RUCC meeting as follows:

• ► Review current situation.
• ► Decide how to make progress of this TF-RUCC activity
• ► Develop working plan with time schedule.

TF members agreed on that.

Draft minutes were finalized without any modifications

Draft minutes were finalized without any modifications

Review of the development of Flex-GTR corridors during the previous phases of development of the pedestrian safety GTR (through the Technical Evaluation Group (TEG)).

Chairperson presented TR-RUCC-K-04.

Chairperson: we should check current GTR leg components if they pass the original GT corridors.

Bernard Been: all current legs were developed to pass the agreed TEG GTR component corridors. The GT tests and corridors are no longer relevant. If the GTR does not pass the GT corridors, could it mean that all 23 legs need bone replacement? What is the purpose of repeating GT tests? The group should focus on dynamic test corridors.

Chairperson: we expect only a small influence of tests condition change for component level certification tests, but we have not checked this. It is important to review all steps of the certification procedure. For component check (PE sheet to roller carriage comparison) 3 bones would be sufficient.

Bernard Been: bone material is missing in the slides as a variable. Would be good to show in the slides.

Mark Burleigh: we did not see major problems with first Vinyl ester bone batch, problems only showed up with subsequent batches.

Oliver Zander: would be good to show number of tests used to set corridors and number of legs.

Mark Burleigh: asked for further testing to be added to future action plan we should retest due to different bone material, different bone batches and longer rubber flesh. The component review would be needed first.

Oskar Ries: If legs go into regulation we must have reliable corridors that can be met in the future. The group must be confident that the legs are build consistently and reproducible, before it can commit to spending a big effort on a large test matrix. As starting point, we must first establish that the legs going into the round robin meet the current GTR component tests consistently.

Chairperson: Based on our discussion at this meeting, TR-RUCC-K-03 and TR-RUCC-K-04 will be modified (TR-RUCC-K-03-Rev.1 and TR-RUCC-K-04-Rev.1). As for the activity schedule after Jan. 2012, we will discuss again at the 1st TF-RUCC meeting in Jan. 2012.

Updated version of the historical review of the development of the certification corridors for the Flex-GTR test tool developed for the phase 2 amendments to GTR 9 on pedestrian safety.

As chair of the task force, Dr. Konosu presented the revision 1 of document GTR9-1-03 containing the Terms of Reference for the TF-RUCC. He also presented document TF-RUCC-K-04-Rev.1 regarding the derivation of certification corridors and on the future action plan of the TF. He finalized that the group would like to finish its work by end of March 2012 in time for the 2nd meeting of the informal group.

Mr. Hohmann mentioned that the time schedule seems very tough. Mr. Been replied that it is tough but not impossible to achieve and that Humanetics would need the support of this group, specifically regarding the availability of impactors and maybe also regarding test capacity. Currently, five prototype legforms and 22 serial production legforms exist; additional 12 legforms currently are ordered but not yet delivered to the customers. Mr. Hohmann stressed that, however, Humanetics first need to provide some impactors with an identical build level to produce new data and to assure that the data are comparable. With these data, discussion on the certification corridors can be restarted but keeping in mind that industry would refuse the corridors being widened up. Dr. Ries underlined again that from the OEM’s point of view the corridors need to be as tight as possible but that shifting the corridors could be acceptable. The chair of the task force promised that the comments above will be considered accordingly.

Mr. Been replied that Humanetics first needs an agreement that the corridors can be redefined. With the latest modifications to the impactor it is impossible to meet the certification corridors and that data shown by Humanetics from different tests underline this. He believes that different labs should provide their data from which new corridors could be derived. The chair of the informal group added that the corridors established by the TEG should serve as starting point for further analysis.

Mr. Zander added that the very first step needs to be tests being carried out in one experienced test lab with a limited number of legforms (later on it was clarified that preferably the legforms should be serial production legforms with bone core material from different batches, as this will also occur in the future) to assure repeatability and reproducibility of the test results but excluding lab to lab variability. Starting form this point, involvement of further labs etc. can go on. Dr. Ries added that the production process of the FlexPLI must guarantee that impactors which will be sold in future will behave like those ones on which the possible new corridors will be derived from. Mr. Been replied that

Humanetics will do their very best to do so but that sometimes changes are necessary due to circumstances that cannot be controlled by Humanetics, e.g. material availability. The chair added that this of course can happen. However, it needs to be well documented and then reported to a forum like this group. Therefore detailed information and documentation on the build levels of the impactors is a prerequisite. The respective forum then may decide to e.g. revise test procedures, corridors etc. Mr. Zander underlined the link of the different issues like repeatability, impactor output and possible shift of certification corridors. A changed performance would not only mean a possible shift of certification corridors but also a possible modification of the impactor threshold values. Mr. Hohmann requested that all changes to the legform need to be well documented. After some discussion on this Mr. Been stated to bring back all this information to his company’s quality assurance group and that he will report on this.

Mr. Knotz mentioned that the informal group should also discuss the speed measurement for the FlexPLI: Currently, there is no procedure foreseen and this is also a potential issue.

The chair wondered whether the PADI or the users’ manual should cover such details. Humanetics will check if this can be incorporated.

After some discussion on all the details mentioned above it was agreed again that the task force will consider the discussion in this group in their future work.

Chairperson presented TR-RUCC-K-04.

Chairperson: we should check current GTR leg components if they pass the original GT corridors.

Bernard Been: all current legs were developed to pass the agreed TEG GTR component corridors. The GT tests and corridors are no longer relevant. If the GTR does not pass the GT corridors, could it mean that all 23 legs need bone replacement? What is the purpose of repeating GT tests? The group should focus on dynamic test corridors.

Chairperson: we expect only a small influence of tests condition change for component level certification tests, but we have not checked this. It is important to review all steps of the certification procedure. For component check (PE sheet to roller carriage comparison) 3 bones would be sufficient.

Bernard Been: bone material is missing in the slides as a variable. Would be good to show in the slides.

Mark Burleigh: we did not see major problems with first Vinyl ester bone batch, problems only showed up with subsequent batches.

Oliver Zander: would be good to show number of tests used to set corridors and number of legs.

Mark Burleigh: asked for further testing to be added to future action plan we should retest due to different bone material, different bone batches and longer rubber flesh. The component review would be needed first.

Oskar Ries: If legs go into regulation we must have reliable corridors that can be met in the future. The group must be confident that the legs are build consistently and reproducible, before it can commit to spending a big effort on a large test matrix. As starting point, we must first establish that the legs going into the round robin meet the current GTR component tests consistently.

Chairperson: Based on our discussion at this meeting, TR-RUCC-K-03 and TR-RUCC-K-04 will be modified (TR-RUCC-K-03-Rev.1 and TR-RUCC-K-04-Rev.1). As for the activity schedule after Jan. 2012, we will discuss again at the 1st TF-RUCC meeting in Jan. 2012.

Chairperson will make report to informal group.

The intention of this meeting is to provide an understanding to TF members of the calibration procedures. Technical discussions will start in January.

This TF group is dedicated to certification corridors only. If different technical problems occur, IG GTR9-PH2 decides whether new TFs will be formed or not.

Chairperson (Atsuhiro Konosu) presented terms of reference for TF-RUCC-K-03. All GTR certification procedures are to be reviewed and to be updated if needed.

Bernard Been: In the current schedule there is no activity of Round Robin testing. We do not want to make the same mistake (rush to make certification corridor with limited data) and must carry out Round Robin, test 10 legs with current material batches from at least 5 labs. Oskar Ries supported this statement.

It was asked when 10 legs could be made available. This would depend on customer co operation when they can they spare legs for testing.

Bernard Been: it is possible to complete round robin in 21 weeks if well managed allowing for transportation and efficient testing, e.g 4 legs at a time.

Carsten Hohmann: how many legs have been made? Mark Burleigh answered 23 customer legs have been built, and 5 prototypes (3 JARI, 2 Humanetics).

Round robin legs should be certified (component and assembly level) at Humanetics before Round Robin testing to ensure consistent spec is present in all legs.

Oliver Zander: can we look at old data and not carry out expensive testing?

Steve Pingston: do we have agreement to shift corridors?

Chairperson: technical details to be discussed later, component level corridors need to be reviewed first.

What are we to achieve by end of March finalization or draft corridors?

Steve Pingston: what is the goal of this task force?

Dr Konosu: The goal of the RUCC is to make desirable corridors and go step by step with confident data, first focus on component level (tibia, femur, knee), then full assembly by end of March if possible.

Adopted minutes from the first technical session of the task force evaluating the certification corridors for the Flex-GTR test tool developed for the GTR on pedestrian safety.

M Burleigh and K Bambach presented a revision of TF-RUCC-1-03 (TF-RUCC-1-03-Rev.1) to inform their findings as well as their proposals.

Slide 4: It was confirmed that inverse corridors were set with the shorter rubber flesh. BGS and BASt insisted that due to the results shown with the longer flesh that the inverse test was more sensitive to leg changes than the pendulum. The inverse test provides more degrees of freedom than the pendulum test because the impactor is completely released during the impact, while the impactor movement is limited due to its fixations at two positions during the pendulum test.

Slides 5 & 6: There were a number of comments regarding the short rubber flesh data. BASt requested the test results, the impactors that have been used, the check up procedure etc. K Bambach is to send out a spread sheet to explain calculations. Inverse values were low and Humanetics will rerun these tests with closer values to the corridors. Chairperson recommended to Humanetics that to review and update Humanetics inverse test rig because quite low data is very strange. It has a chance that a test rig problem.

Slide 7: Humanetics proposed the use of PE sheet to calibrate component sub assemblies as this setup proved to be more stable and repeatable; BASt/BGS would prefer gap condition roller set up as it is more sensitive. M Burleigh stated the setup is very difficult and open to inconsistent results. The fixture has been seen to slide on the top of the roller carriages and rollers can slide rather then roll even when a gap condition is present. The nature of the rollers is very unstable which is why there are stop plates around the rollers. The high movement of the carriage adds to this instability.

The gage sensitivity is still calculated on rollers as the lower travel makes this test more stable than the assembly. Chairperson supported Humanetics proposal to use PE sheet to calibrate component sub assemblies because roller set up is significantly not userfriendly and difficult to handle that to obtain appropriate test data.

Slide 8: BASt and BGS stated the “no gap” test being an unusual calibration setup because it causes unintended friction.

Slide 10: BASt and BGS confirmed this being the usual test setup for e.g. the calibration of load cells.

Slide 11: The conclusion is logical because, as stated by BGS and BASt, the “no gap” test setup causes unintended friction.

Slide 14: BGS and BASt gave an explanation for the observations made during the tests within TF-RUCC-1-05-Rev1: The “with gap” test setup is more sensitive and thus avoids unintended friction, provides a higher degree of freedom, a higher influence on the long bone properties, higher values and a better assessment. Therefore, the higher repeatability of the “PE test” results should not be used as an argument for choosing this type of test, but just the other way round the PE test and the “no gap” test are considered to not being sensitive enough. Therefore, it is again proposed to use the “with gap” test as a calibration test.

Slide 15: Subsequent to the previous comments made, BASt and BGS requested this comparison using the “with gap” method.

Slides 20-21: There was concern over the removal of the outliners for sensitivity calculation on batch 3 and 4 bones. Humanetics had left in the calculation with outliners as that had been previously shown in the earlier presentation (TF-RUCC-1-03) and to help explain the change they were left in. The justification to remove the outliners was; normally bones are flexed 4 times before running the test, in these two cases (by mistake) they were not.

Slide 22: Humanetics stated existing legs would be used for the round robin set up. These legs will have new bones, meniscus, knee springs, bone interfaces, knee cables, bone cables, flesh system and rubber segment buffers. All other parts will be assessed for wear/damage and replaced as needed. It was not possible to supply new legs. It was confirmed that any production legs being made in the assessment time period would have 5 pendulum and 5 inverse tests to add to the data. BASt repeated their concerns that again no brand-new legforms being used to possibly establish shifted or modified corridors, if needed.

Slide 23: J C Kolb confirmed Bertrandt are prepared to run round robin assessment and M Chaka from Ford also offered their services for round robin. O Zander requested that before any round robin testing they need to test two legs to the latest build level to check repeatability and reproducibility as discussed in the IG meeting in Geneva, then round robin should go ahead if results prove satisfactory. Kurt Bambach offered to send Humanetics data showing repeatability on dynamic testing. BASt wondered why the pendulum test being less documented than the inverse test and requested an identical documentation, where applicable.

Slide 24: It was stated that Humanetics would recalibrate the string pots before the round robin to ensure they were to specification. D Gerhring asked if this was necessary. The pots should be calibrated annually anyway and it would be prudent to do this before testing.

M Burleigh and K Bambach presented a revision of TF-RUCC-1-03 (TF-RUCC-1-03-Rev.1) to inform their findings as well as their proposals.

Slide 4: It was confirmed that inverse corridors were set with the shorter rubber flesh. BGS and BASt insisted that due to the results shown with the longer flesh that the inverse test was more sensitive to leg changes than the pendulum. The inverse test provides more degrees of freedom than the pendulum test because the impactor is completely released during the impact, while the impactor movement is limited due to its fixations at two positions during the pendulum test.

Slides 5 & 6: There were a number of comments regarding the short rubber flesh data. BASt requested the test results, the impactors that have been used, the check up procedure etc. K Bambach is to send out a spread sheet to explain calculations. Inverse values were low and Humanetics will rerun these tests with closer values to the corridors. Chairperson recommended to Humanetics that to review and update Humanetics inverse test rig because quite low data is very strange. It has a chance that a test rig problem.

Slide 7: Humanetics proposed the use of PE sheet to calibrate component sub assemblies as this setup proved to be more stable and repeatable; BASt/BGS would prefer gap condition roller set up as it is more sensitive. M Burleigh stated the setup is very difficult and open to inconsistent results. The fixture has been seen to slide on the top of the roller carriages and rollers can slide rather then roll even when a gap condition is present. The nature of the rollers is very unstable which is why there are stop plates around the rollers. The high movement of the carriage adds to this instability.

The gage sensitivity is still calculated on rollers as the lower travel makes this test more stable than the assembly. Chairperson supported Humanetics proposal to use PE sheet to calibrate component sub assemblies because roller set up is significantly not userfriendly and difficult to handle that to obtain appropriate test data.

Slide 8: BASt and BGS stated the “no gap” test being an unusual calibration setup because it causes unintended friction.

Slide 10: BASt and BGS confirmed this being the usual test setup for e.g. the calibration of load cells.

Slide 11: The conclusion is logical because, as stated by BGS and BASt, the “no gap” test setup causes unintended friction.

Slide 14: BGS and BASt gave an explanation for the observations made during the tests within TF-RUCC-1-05-Rev1: The “with gap” test setup is more sensitive and thus avoids unintended friction, provides a higher degree of freedom, a higher influence on the long bone properties, higher values and a better assessment. Therefore, the higher repeatability of the “PE test” results should not be used as an argument for choosing this type of test, but just the other way round the PE test and the “no gap” test are considered to not being sensitive enough. Therefore, it is again proposed to use the “with gap” test as a calibration test.

Slide 15: Subsequent to the previous comments made, BASt and BGS requested this comparison using the “with gap” method.

Slides 20-21: There was concern over the removal of the outliners for sensitivity calculation on batch 3 and 4 bones. Humanetics had left in the calculation with outliners as that had been previously shown in the earlier presentation (TF-RUCC-1-03) and to help explain the change they were left in. The justification to remove the outliners was; normally bones are flexed 4 times before running the test, in these two cases (by mistake) they were not.

Slide 22: Humanetics stated existing legs would be used for the round robin set up. These legs will have new bones, meniscus, knee springs, bone interfaces, knee cables, bone cables, flesh system and rubber segment buffers. All other parts will be assessed for wear/damage and replaced as needed. It was not possible to supply new legs. It was confirmed that any production legs being made in the assessment time period would have 5 pendulum and 5 inverse tests to add to the data. BASt repeated their concerns that again no brand-new legforms being used to possibly establish shifted or modified corridors, if needed.

Slide 23: J C Kolb confirmed Bertrandt are prepared to run round robin assessment and M Chaka from Ford also offered their services for round robin. O Zander requested that before any round robin testing they need to test two legs to the latest build level to check repeatability and reproducibility as discussed in the IG meeting in Geneva, then round robin should go ahead if results prove satisfactory. Kurt Bambach offered to send Humanetics data showing repeatability on dynamic testing. BASt wondered why the pendulum test being less documented than the inverse test and requested an identical documentation, where applicable.

Slide 24: It was stated that Humanetics would recalibrate the string pots before the round robin to ensure they were to specification. D Gerhring asked if this was necessary. The pots should be calibrated annually anyway and it would be prudent to do this before testing.

A Konosu/chairperson presented TF-RUCC-1-04 to introduce Japan observations as well as Japan proposals. Proposal was to re-examine all the GTR corridors by preparing bones that are in the middle of the bone corridor to avoid going back to discussion on injury thresholds. Japan proposed to carry out this testing at JARI because JARI has the original test rigs which were used to make static and pendulum corridors.

Different results were obtained with bone core between JARI and Humanetics. M Burleigh requested the JARI bone fixture to set up the round robin bones and look to see if it was the fixture or the test set up that was the problem.

JARI propose the use of the PE sheet as opposed to the rollers as it is more stable, repeatable and easier to control.

BASt again opposed that a high repeatability in that context should not alone decide on whether to use the PE sheet test setup because the repeatability is due to unintended friction.

Dr Ries confirmed the importance of a stable consistent build, M Burleigh said this proposed procedure from the bone core up should ensure this.

O Zander showed comments to slides 22 and 23 of the Humanetics presentation. He strongly recommended that SN02 was not used in the round robin as it was the only leg left of the original three prototypes and would be the only reference to the original testing. Besides, the leg has seen many tests and has some damage. It was agreed not to use SN02 in the round robin and it is to be left untouched unless there is a request for repair from the owner JAMA.

O Zander showed comments to slides 22 and 23 of the Humanetics presentation. He strongly recommended that SN02 was not used in the round robin as it was the only leg left of the original three prototypes and would be the only reference to the original testing. Besides, the leg has seen many tests and has some damage. It was agreed not to use SN02 in the round robin and it is to be left untouched unless there is a request for repair from the owner JAMA.