Download ppt - Marco Anghileri Dipartimento di Ingegneria Aerospaziale Politecnico di Milano Italy

Marco AnghileriDipartimento di Ingegneria Aerospaziale

Politecnico di MilanoItaly

Robust: “Road Upgrade of Standards”

GRD1-2002-70021Second year meeting.

Copenhagen March 1 – 2 2005

WP3: “Instrumentation and measurement”

2 /35 Second year. Copenhagen 1/3/2005

Robust

. G

RD

1-2

002-7

002

1

2

Composite mounting block activity report.

• Composite mounting design and construction.

• Test at TRL.• Tests at Polimi.• Discussion


Robust

. G

RD

1-2

002-7

002

1

3

Scope of this activity

• Acceleration recorded on CG contains components introduced by the oscillation of the floor.

• The structure of the accelerometers mounting block can modify these frequencies shifting them at lower regions where our severity indices work.

• To investigate the influence on the acceleration time histories of mounting block mass, a carbon fiber mounting block has been designed constructed and used during a TB11 test and several deceleration tests.


Robust

. G

RD

1-2

002-7

002

1

4

Composite mounting block

• Requirements:– Weight (about 400 g).– Shape (similar to Lier structure)– Strength (able to sustain several crash

test).

• 10 blocks produced.


Robust

. G

RD

1-2

002-7

002

1

5

TRL TB11 test

• TRL performed a Round Robin II test installing the new structure.

• The fixing to the floor structure was obtained with 4 bolts

upper plate to fix accelerometersWhite Fiberglass plates

lower fiberglass plate to fix the structure to the car

4 Bolts

Car floor

Aluminium or steel plate


Robust

. G

RD

1-2

002-7

002

1

6

TRL TB11 test

• Test was performed at TRL on Nov 17 2004.

– Ford Fiesta:– Weigth 900 kg.– Impact angle 20°– Speed 102 km/h.


Robust

. G

RD

1-2

002-7

002

1

7

TRL TB11 test results.

• Structural behaviour:– The structure of the mounting did not show any

damage during the test.

• Frequency behaviour:– Not satisfactory. Presence of oscillation.

• A first numerical acitivty has been conducted to investigate these problems


Robust

. G

RD

1-2

002-7

002

1

8

TRL TB11 results

• Frequency analysis:– During the test two set of accelerometers have

been mounted on the car. The first on the new mounting and the second on the floor of the car.

0 50 100 150 200 250 300 350 400 450 5000

1

2

3

4

5

x 104 spectrum power density

Tunnel composite


Robust

. G

RD

1-2

002-7

002

1

9

TRL TB11 results– The frequency content at lower frequency (below 20 hz) is lower

for the new mounting than for the second set of accelerometers.– The frequency content at higher frequency is lower for the new

mounting than for the second set of accelerometers.– The new mounting block shows a strong resonance peak at 50 hz

(any contribution from AC supply has been excluded).

0 50 100 150 200 250 300 350 400 450 5000

1

2

3

4

5

x 104 spectrum power density

Tunnel composite


Robust

. G

RD

1-2

002-7

002

1

10

TRL TB11 results• Time analisys:

• Time analysis shows that acceleration measured on the new mounting are always lower than the acceleration measured on the second set. Different accelerometers location should be taken into account. This is clearly shown in the following figures with unfiltered results:

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-120

-100

-80

-60

-40

-20

0

20

40

60

80acceleration x

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-120

-100

-80

-60

-40

-20

0

20

40

60

80acceleration x tunnel

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-120

-100

-80

-60

-40

-20

0

20

40

60

80acceleration x composite


Robust

. G

RD

1-2

002-7

002

1

11

TRL TB11 results.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration y

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration y tunnel

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration y composite

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration z

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration z tunnel

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8-200

-150

-100

-50

0

50

100

150acceleration z composite


Robust

. G

RD

1-2

002-7

002

1

12

TRL TB11 results

• Same behaviour if a 20 hz filtering is applied.

– Accelerations time history measured on the tunnel does not show the typical second peak related to the impact of the rear part of the vehicle while composite acceleration shows this second impact. A correction should be applied for the different location.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9-20

-15

-10

-5

0

5acceleration x

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9-35

-30

-25

-20

-15

-10

-5

0

5acceleration y

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9-6

-4

-2

0

2

4

6

8

10

12acceleration z


Robust

. G

RD

1-2

002-7

002

1

13

TRL TB11 results

• Severity indices

0.1 0.15 0.2 0.25 0.3 0.35 0.40

0.5

1

1.5

2

2.5

ASI Thiv Phd

Tunnel 2.75 38.7 (t=.2531) 11.7

Composite 1.97 24.3 (t=.2883) 31.2

– The standard correction should be applied for the different location.


Robust

. G

RD

1-2

002-7

002

1

14

TRL TB11 conclusion.

• The previous results show interesting aspects that should be further investigated.

• Preliminary analysis showed a basic success but the mechanical noise affecting acceleration measurement must be investigated and eliminated.

• A new activity was planned at Polimi to investigate this problem.


Robust

. G

RD

1-2

002-7

002

1

15

Composite mounting fixing.

• To better understand the results of the test carried at TRL on November 17 2004 some basic experiments has been performed.

• The composite mounting block has been installed on a Peugeot 106 tunnel to reproduce the resonance problems shown by TRL data.


Robust

. G

RD

1-2

002-7

002

1

16


– A first series of tests have been conducted using the installation procedures agreed by the consortium. The structure has been fixed on the tunnel of the vehicle using 4 bolts (8 mm diameter).

– To measure the frequency behaviour a standard technique for modal analysis has been adopted impacting the structure with a small hammer and measuring the output with piezoelectric accelerometers with high sensitivity


Robust

. G

RD

1-2

002-7

002

1

17



Robust

. G

RD

1-2

002-7

002

1

18



Robust

. G

RD

1-2

002-7

002

1

19


• A triaxial accelerometer has been installed on the structure.

• The output of these accelerometers, together with the spectrum response showed a behaviour similar to the one seen at TRL even if the main frequency is here at about 85 hz.

• The difference between this frequency and the one measured at TRL (50 hz) is probably due to the different placement of the structure and the presence, at TRL, of the heavy structure used to release the cable that lowers all the natural frequency increasing the mass of the adjacent structure.


Robust

. G

RD

1-2

002-7

002

1

20

0 20 40 60 80 100 120 140 160 180 2000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1spectrum power density

Composite mounting fixing.• This figures show these important results:

» The behaviour shown during TRL tests has been reproduced even if the resonance frequency is at 85 hz

» This is the only frequency present and if we are able to shift this frequency where severity indices don’t work all the possible concerns related with these oscillation are removed.

0 0.5 1 1.5 2 2.5 3

x 104

-5

-4

-3

-2

-1

0

1

2

3

4

5acc 1acc 2acc 3

0 20 40 60 80 100 120 140 160 180 2000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9



Robust

. G

RD

1-2

002-7

002

1

21

Composite mounting new fixing.

– To overcome the above shown problem we simply added four more bolts and introduced spacers to fix the structure on a wider part of the floor.


Robust

. G

RD

1-2

002-7

002

1

22

Composite mounting new fixing.– The first natural frequency of the system is above

105 hz (y direction) ad 140 Hz (x direction) . These frequencies are far from the relevant for the severity indices evaluation.

0 0.5 1 1.5 2 2.5 3

x 104

-3

-2

-1

0

1

2

3acc 1acc 2acc 3

0 20 40 60 80 100 120 140 160 180 2000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9


0 20 40 60 80 100 120 140 160 180 2000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9



Robust

. G

RD

1-2

002-7

002

1

23

Deceleration tests.

• Having solved the problems related to the mechanical oscillation a new activity started installing the car on the deceleration sled at Polimi.

• Purpose:– To apply known deceleration triangular pulses

comparing measures obtained with the composite structure and an aluminium structure.

– Same pulse to different mounting structures.– To have results avoiding any cotribution from

barrier, ground condition or suspension system.


Robust

. G

RD

1-2

002-7

002

1

24

Deceleration tests.

• Aluminium mounting block structure:– Simple shape, same dimensions as composite

structure.– Weigth: 4.7 kg

• Two series of tests. – Car installed with two directions respect to the

applied acceleration.

70°20°

Imposed acceleration

Car

Sled


Robust

. G

RD

1-2

002-7

002

1

25

Deceleration tests.


Robust

. G

RD

1-2

002-7

002

1

26

Deceleration tests.


Robust

. G

RD

1-2

002-7

002

1

27

Deceleration tests

• Performed 12 tests:– 3 tests 20° aluminium mounting– 3 tests 20° composite mounting– 3 tests 70° aluminium mounting– 3 tests 70° composite mounting

• Acceleration pulses:– Peaks: 5 to 25 g (imposed ASI from .45 to 1.9)– Rise time: 0.1 to 0.15 s.

• Data analisys not completed but interesting results.


Robust

. G

RD

1-2

002-7

002

1

28

Deceleration tests


Robust

. G

RD

1-2

002-7

002

1

29

Deceleration tests. Results

• Imposed acceleration time histories properly reconstructed by local measures.

1.6 1.8 2 2.2 2.4 2.6-25

-20

-15

-10

-5

0

5

10

15

20

25Aluminium mounting. Comparison. Unfiltered

Car Sled

2 2.2 2.4 2.6 2.8 3-20

-15

-10

-5

0

5

10

15

20

25Composite mounting. Comparison. Unfiltered

Car Sled


Robust

. G

RD

1-2

002-7

002

1

30

Deceleration tests. Results.

• Frequency response:– Y direction much more sensible (70° impact).– Differences also at lower frequencies

0 20 40 60 80 1000

1000

2000

3000

4000

5000

6000

7000

8000spectrum power density. y direction. 70° impact

AluminiumComposite

2 4 6 8 10 12 140

1000

2000

3000

4000

5000

6000

7000

spectrum power density. y direction. 70° impact

AluminiumComposite


Robust

. G

RD

1-2

002-7

002

1

31

Deceleration tests. Results.

• Accelerations measured on composite mounting block have always:– lower oscillation amplitude– higher natural frequency

2.5 3 3.5 4 4.5

x 104

-20

-15

-10

-5

0

5

10

15

20

Car accelration y direction. Unfiltered. Composite

2 2.5 3 3.5 4 4.5

x 104

-20

-15

-10

-5

0

5

10

15

20

Car accelration y direction. Unfiltered. Aluminium


Robust

. G

RD

1-2

002-7

002

1

32

Severity indices. ASI– No differences during first part of the deceleration

pulse due to the strong slope of the acceleration pulse.

– After the peak differences between mounting blocks is relevant.

– Better agreement between sled and mounting with composite solution.

1.6 1.8 2 2.2 2.4 2.6-25

-20

-15

-10

-5

0

5

10

15

20

25Aluminium mounting. Comparison. Unfiltered

Car Sled

2 2.2 2.4 2.6 2.8 3-20

-15

-10

-5

0

5

10

15

20

25Composite mounting. Comparison. Unfiltered

Car Sled


Robust

. G

RD

1-2

002-7

002

1

33

Severity indices. ASI

• Max difference measured between ASI measures is 0.1

1.95 2 2.05 2.1 2.15 2.2

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8ASI Traditional Comparison. Aluminium

Car Sled

2.45 2.5 2.55 2.6 2.650

0.1

0.2

0.3

0.4

0.5

0.6

0.7

ASI Traditional Comparison. Composite

Car Sled


Robust

. G

RD

1-2

002-7

002

1

34

Severity indices. ASI

• Less differences with filtered ASI

2.45 2.5 2.55 2.6 2.65 2.70

0.1

0.2

0.3

0.4

0.5

0.6

0.7

ASI Filtered Comparison.Composite

Car Sled

1.95 2 2.05 2.1 2.15

0.1

0.2

0.3

0.4

0.5

0.6

0.7

ASI Filtered Comparison. Aluminium

Car Sled


Robust

. G

RD

1-2

002-7

002

1

35

Composite mounting block. Conclusion

• The activity showed the influence of the mass of the mounting block on the results.

• Ferquency response can be modified with a light structure.

• The composite solution could be not the proper solution but some specification on the frequecy response should be added to EN 1317.

• Autostrade will test next week the N2 barrier with the composite block