New ISO for H3 Chest Displacement

[Josep Maria Pons]

Hello all,

As EuroNCAP TWG decided in their meeting on last April 23rd, two ways of getting Chest displacement for H3 dummies will be possible:
1) the standard calculation based on existing lineal (1st degree) potentiometer calibration.
2) the new calculation based on (3rd degree) potentiometer calibration, wich offers some improvement in chest displacement results.

This new calculation is required by EuroNCAP for frontal crash tests for H3 dummies (driver and passenger).
Taking into account that the rest of Regulations are using the first option, I suggest a new ISO code is to be generated in order to
distinguish between the two cases.

In IDIADA, we are using the next new code for the new option: ??CHST0003H3DSX?
and the already known code for the first option : ??CHST0000H3DSX?
This is only used in IDIADA meanwhile there is no official ISO code decided for this option.

I'm waiting for your opinion about this question.

Best Regards

Josep Maria Pons

[DiVe]

Dear Mr Pons,
there has already been some discussion on this issue in the working group. I would agree that we need a new coding to distinguish between the "old" method chest deflection and the new variant using the 3rd order polynom.
Actually the processing will be done with the unfiltered voltage signal but we expect that the polynom will act like a filtering on the signal. Therefore we might also define a special filter setting in the code to may avoid further filter processing for this channel ("3" is not available!). I am not quite sure if Euro NCAP processing still requires filtering with CFC180 after the polynom processing. Actually I have no example channel available to decide whether it makes sense or not to apply the CFC180 filter.
If some other partner has already practical experience or results it will be helpful to know about this.

Nevertheless I personally would also agree that it might be better signaled that there is a "different" channel by also using modified coding in the fine loactions. Especially as the Fine Location 1 and 2 are free in the potentiometer coding.

To have some more options for the discussion:
??CHST0003H3DSX?
??CHST0003H3DSXS -> "S" hase been decided for the V1.5 of the database as "placeholder" for "special filtering"
??CHST00P3H3DSXS -> P3 = Polynom 3rd order, but may confusing as P3 is already in FL3 available and also FL2 needs extension for "P3"

CO in FL2 like "Corrected" is not available as "CO" is already used for "Compression"

I personally could imagine to use the:
??CHST0003H3DSXS
So there is nothing new needed in the specific parts of the code and the difference is not just hidden in the Filter Class part of the code.


Kind regards,

Dirk Vetter

[derpmann]

Hello,
when we discussed this issue in our company a few weeks ago we saw the necessity to exchange the data of the same channel two times.
For EuroNCAP with polynom calibration and for all other regulations with the old calibration method.

We prefer the storage and exchange of only one channel with the known Channel Code ??CHST0000H3DSX?
and the following constraints:

• The values of sample are the balanced values computed with the old calibration method.
• The header of the channel file is expanded by 7 new descriptors which allow the computation of the channel data according to the EuroNCAP specification.

The new descriptors are optional as a complete group, but mandatory whenever one of them is used.
This behaviour is easier to handle with the blocking mechanism of MME V2.0 but nevertheless possible with MME V1.x.
The values of the new descriptors specify the offset correction and the used sensitivity value for the old calibration procedure and
the 4 coefficients (including the offset = coefficient of 0th order) of the polynomial procedure.
These are in detail:

Code: Select all

Inverse sensitivity        :                            Float [Unit/Sensor output]
Inverse polynom coeff A    :                            Float [Unit/(Sensor output)^3]
Inverse polynom coeff B    :                            Float [Unit/(Sensor output)^2]
Inverse polynom coeff C    :                            Float [Unit/(Sensor output)]
Inverse polynom coeff M    :                            Float [Unit]
Offset pre test            :                            Float [Sensor output]
Offset post test           :                            Float [Unit]

Procedure:
To the values of sample the "Offset post test" has to be added. The next step is the division by the "Inverse sensitivity" value.
By adding the "Offset pre test" the sensor output can be calculated.
To compute the EuroNCAP values of sample the polynomial calibration can be perfomed using the 4 inverse polynom coefficients.

The advantages of this solution are:

• Most of existing software tools don't have to be changed.
• This solution is portable, if in the future other transducer channels have to be calibrated with the new procedure.
• It is not necessary to create a new channel whenever the polynom procedure has to be used.

Best regards
Peter Derpmann-Hagenström

[derpmann]

Recommendation of the MME TaskForce:

• For the data exchange only the channel file with the linear calibrated data has to be used.

• The values necessary for the conversion from the linear to the polynomial calibrated data have to be exchanged with the aid of some new descriptors of the channel header.

• The unit used for the coefficients has to be equal to the unit of the data values of the linear calibrated channel.

• If it is necessary for storage or illustration to use both kinds of calibration at the same time, the polynomial calibrated data has to be marked by the value 03 in the FineLocation2 of the channel code.

This recommendation will be integrated into MME 1.6.
For some more information see the attached pdf-document from 08.06.2009.

Best regards
Peter Derpmann-Hagenstroem

[derpmann]

Integrated in RED A 1.6 chapter A.2.2.9 .