Global Nuclear Fuel

Verification of LS-DYNA Finite Element Impact Analysis by Comparison to Test Data and Classic First Principle Calculations
Andy Langston and Victor Smith
PATRAM 2010 3-8 October, 2010, London, England, UK

Introduction
Development:
RAJ-II BWR fresh fuel transport package developed in Japan as replacement for first generation design. Drop tested in Japan for METI certification. Drop tested in USA at Oak Ridge, TN facility for NRC SAR. Drop tested in France by Japanese to validate loose rod container.

Licensing History:
Licensed in Japan in the mid 1990s. Licensed in USA in 2005 as replacement for GNF and AREVA first generation packages. 2007 Present, GNF and Westinghouse licensing package in EU. During the licensing review additional information requested concerning the impact performance of the package with respect to IAEA TS-R-1.

LS-DYNA Model
Solid Model:
Solid model developed in AutoDesk Inventor. Model developed from fabrication drawings. Crushable materials modeled as solid objects. Sheet metal modeled as Surfaces.

LS-DYNA Model
FEA Model:
Solid model imported into ANSYS Workbench.
Package Assembly

Model meshed with Workbench meshing tools. LS-DYNA keyword file created In ANSYS Mechanical. Crushable materials modeled with solid elements. Sheet metal modeled with shell elements.
Space Frame

Presentation3

Inner Container with Honeycomb Block

Total of 534853 nodes and 442331 elements

Rigid Fuel Bundle Payload

LS-DYNA Model
Stress (psi)

250

200

150

100

Honeycomb Engineering StressStrain Properties

Nominal (21C) Hot (77C) Cold (-40C)

Material Properties:
Honeycomb and Ethafoam properties obtained through laboratory testing. Three temperature ranges test including 40C, 21C, and 77C that represents cold, ambient, and hot conditions. LS-DYNA material types *MAT_HONEYCOMB and *MAT_CRUSHABLE_FOAM
Stress (psi)
100 90 80 70 60 50 40 30 20 10 0 0.000

50

0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

90000 80000 70000

0.80

0.90

Strain (in/in)

Total True Stress (psi)

60000 50000 40000 30000 20000 10000 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

True Stress Versus True Strain for 304 SS

Plastic True Strain (in/in)

Nominal (21C) Hot (77C) Cold (-40C)

Ethafoam Engineering Stress-Strain Properties.

0.020 0.040 0.060 0.080 Strain (in/in) 0.100 0.120 0.140 0.160

LS-DYNA Model
Benchmark with Test Results:
LS-DYNA honeycomb material property defines an instantaneous modulus of elasticity that accounts for the column buckling of the honeycomb cell. The instantaneous modulus of elasticity was adjusted until the initial peak acceleration matched the French top drop test results. The French drop test represents the best recorded data for any of the RAJII test programs.
200

150

100
Acceleration (g)

50

-50 RAJ-III -100 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 Time (s) 0.0035 0.004 0.0045 0.005 Top Drop 21C

Benchmark of LS-DYNA with Drop Test Results

Analysis Results
Side Drop:
Maximum accelerations occur when lightest fuel bundles is coupled with coldest temperature (-40C). Accelerations increase 5%. Heaviest fuel bundle coupled with hot conditions results in 9% decrease in accelerations. The peak acceleration is 340g at 500 Hz.
400 350 300 250 200

Acceleration (g)

150 100 50 0 -50 -100 0 0.002 0.004 0.006 0.008 0.01 Time (s)
Side Hot Heavy

0.012

0.014

0.016

Side Ambient Nominal

Side Cold Lite

Side Drop Accelerations (Cold, Ambient and Hot Conditions)

Analysis Results

Top Drop:
Acceleration (g)

200

Like the side drop maximum accelerations occur when lightest fuel bundles is coupled with coldest temperature (-40C). The peak acceleration is 186g at 500 Hz.

150

100

50

-50

0.0025

0.005

0.0075

0.01

0.0125 Time (s)

0.015

0.0175

0.02

0.0225

0.025

Top Ambient Nominal

Top Hot Heavy

Top Cold Lite

Top Drop Accelerations (Cold, Ambient and Hot Conditions)

Analysis Results
Slap-down/Whiplash:
Acceleration (g)
160 140 120 100 80 60 40 20 0 -20 0 0.0025 0.005 0.0075 0.01 0.0125 Time (s) Slap Down Top Ambient Nominal 0.015 0.0175 0.02 0.0225 0.025

The analysis results show that the RAJ-II is more efficient during the slap-down event than the flat top drop. During slap-down, honeycomb surface area is initially only available at the point impact and gradually increases as the impact progresses. Due to the geometry of the packaging, the initial peak acceleration is much higher during the flat top or side events.

Top Drop versus Slap-down Accelerations

Benchmarking
Comparison with Historic Test Results:
To benchmark the LS-DYNA analysis results, comparison to historic drop test is used. The RA-3D package is a first generation design similar to the RAJ-II RA-3D drop tests included natural uranium bundles of common designs to perform the regulatory testing.
Comparison of LS-DYNA and RA-3D Test

Good agreement between the RAJ-II LS-DYNA analysis results and RA-3D test results including the impact duration. The peak acceleration of the RA-3D is higher than that of the RAJ-II because of increased honeycomb surface area during the initial impact.

Benchmarking
Impact Predictions with Classic First Principle Calculations:
To further benchmark these results, a hand calculation predicts the peak acceleration.
Acceleration (g)
600 500

400

RA3DPrediction (391g)

300

RAJIIPrediction (229g)
200

Benchmarking possible because of the simple geometry of the RAJ-II and RA-3D honeycomb design. Methodology developed by Mindlin established the basis for predicting acceleration of packaged items.

100

100 0 0.003 0.006

Time (s)

0.009

0.012

0.015

RAJ-II LS-DYNA Side Drop RAJ-II Estimate

RA-3D Side Drop RAJ-II Side Filtered

RA-3D Estimate

LS-DYNA, RA-3D Test, and Hand Calculations

Using this methodology able to provide reasonable estimate.

Verification
Independent Verification using Classic First Principle Calculations:
To further benchmark these results, a hand calculation predicts the peak acceleration. Using the Mindlin method, independent verification was performed to predict the initial and secondary impact during slap-down.

LS-DYNA Slap-Down Results Compared to Hand Calculations

The estimate of the peak acceleration The load was during the initial impact is within 2g of considered to be the acceleration predicted by LS-DYNA. carried by only the The hand calculation estimates a lower cross section of the acceleration for the secondary peak as honeycomb block compared to LS-DYNA. However, the supported by the bottom of the container hand calculation values closely corresponds to the values predicted by parallel to the the computer model. acceleration.

Conclusions
This evaluation shows that testing, finite element analysis, and first principle calculations are all good methods for evaluating the performance of a package. With all methods, the key to good results is having well defined geometry and materials. When all three methods are utilized, it is possible to benchmark analytical models that can be used to further improve packaging design.