As part of the GEOLAB project, the Institute of Geotechnics of TU Darmstadt is calling geotechnical engineers from industry and academia to participate in an international Blind Prediction Contest (BPC) on the response of piles under monotonic and cyclic lateral loading. Two separate tests will be performed on a hollow open-ended steel pile embedded in dry sand. One test under monotonic loading and the other under quasi-static harmonic loading with more than 10,000 loading cycles.
Contestant teams can submit predictions for both tests or for the monotonic test only. The predictions will be objectively marked based on their discrepancy with the experimental values. The teams with the higher score in the prediction of each test will be publicly announced. The rest of the submitted predictions will be anonymised and used for assessing the state of the art and the state of practice by the organiser committee.
An overview of the experimental setup and procedures is provided here.
Please refer to the detailed supporting documentation provided below to prepare your prediction.
Pile test setup
Two tests will be performed at the TU Darmstadt Geotechnical Test Pit. The setup to be used in both tests is schematically depicted in the figure below. It consists in a 3 m hollow steel cylinder embedded 2 m in a dry compacted, very dense medium coarse river sand. In both tests, the pile will be loaded laterally at 0.58 m above the sand level.
Loading stages
The loading sequence for each test will be as follows:
Test 1 (Monotonic)
Stage 1: Controlled displacement at a rate of 1 mm/ min up to a displacement of 20 % of the pile outer diameter (OD) measured at the loading point.
Stage 2: Switch to controlled force, maintaining the achieved force constant for 20 minutes or until the displacement rate is lower than 0.02 mm/ min, whichever occurs last.
Stage 3: Switch to controlled displacement at a rate of -1 mm/ min until the applied force becomes null.
Test 2 (Cyclic)
Stage 1: Controlled displacement at a rate of 1 mm/ min until reaching a reaction force of 20 % of F10 %, where F10 % is the reaction force in Test 1 at a displacement of 10 % of the pile OD measured at the loading point.
Stage 2: 10,000 cycles, with frequency of 0.125 Hz, from 10 % to 30 % of F10 %.
Stage 3: Unloading to zero at a rate of – 1 mm/ min (controlled displacement).
Stage 4: Controlled displacement at a rate of 1 mm/ min until reaching a reaction force of 50 % of F10 %.
Stage 5: Cycling loading, with frequency of 0.125 Hz, from 30 % to 70 % of F10 % until pile failure (see failure definition in the supporting documentation).
Stage 6: Unloading to zero load at a rate of – 1 mm/ min (controlled displacement).
Pile test instrumentation
The pile test will be instrumented as follows:
- 1 force transducer at the connection between the pile and the hydraulic actuator
- 9 displacement transducers conveniently located around the loading point to register X, Y- , Z displacements and rotations.
- 16 strain gauges attached to pile in two longitudinal lines
Sand characterisation tests
The sand has been extensively characterised with laboratory tests, including monotonic (drained and undrained) and cyclic triaxial tests, performed by the Chair of Soil Mechanics, Foundation Engineering and Environmental Geotechnics of Ruhr-University Bochum (Germany), led by Prof. Torsten Wichtmann. In addition, standard Cone Penetration Tests have been performed to characterise the sand condition in depth after compaction.
All the produced data is made available below.
The guidelines of the BPC are given below.
Detailed rules for the contest are made available below.
Contestant teams
The contestant teams may consist of one or more members.
An individual can only be involved in a single team.
Except for category winners, all participants will be kept anonymous. For the publication of the predictions, teams will be identified by an inscription number which will be informed to the team members in advance.
There is no limit in the number of members of a team, but teams with more than one member will designate a corresponding author for communication.
Individuals from TU Darmstadt or other organisations who have witnessed the tests are not allowed to participate.
Teams will be required to fill a template document with a brief description of the methods used for their predictions. All methods (empirical, analytical or numerical) are welcome.
Input data
Detail input data can be downloaded below.
Provided input data includes:
- Detailed description of the test setup and methods.
- Construction drawings and relevant photographs, specifying all necessary geometry and dimensions.
- CPT profiles of the sand pit before testing.
- Sand characterisation tests, including particle size distribution, oedometric compression tests, drained and undrained monotonic triaxial tests (with different confining pressures and sand densities), and cyclic triaxial tests.
Predicted Quantities
The contest will have two categories: “monotonic pile test” and “cyclic pile test”.
All contestant teams must submit a prediction for the monotonic test. Predictions for the cyclic test are optional.
The participants are required to predict the following experimental results:
Monotonic test
Stage 1: Horizontal reaction force, vertical displacements of two points of the pile head, and pile longitudinal strains at given imposed displacements.
Stage 3: Recoverable horizontal displacement after unloading.
Cyclic test (Optional)
Stage 2: For the first cyclic loading package: accumulated vertical and horizontal displacements of two points of the pile head, and pile longitudinal strains after 1, 10, 100, 1000 and 10000 cycles.
Stage 4: For the second cyclic loading package: accumulated vertical and horizontal displacements of two points of the pile head, and pile longitudinal strains after specific 1, 10, 100, and 1000 cycles.
Stage 4: Total number of cycles in the second cyclic loading package.
- The quality of the predictions in each category (monotonic and cyclic tests) will be quantified as a single scalar value, ranging from 0 (the lowest possible quality) to 100 (the highest possible quality).
- The expressions used to obtain theses scores are embedded in the Excel file provided for submitting the predictions (GEOLAB-BPC_00-Prediction&ScoringSheet), specifically in the worksheets “Scoring-Monotonic test” and “Scoring-Cyclic test.”
- For each considered quantity, the error is defined as the absolute value of the difference between the measured value and the predicted value. To allow comparison of the error in different variables, the partial prediction errors are adimensionalised and scaled. Moreover, different weighing factors are applied to displacement and strain errors.
- The prediction quality is evaluated separately for monotonic and cyclical tests, with a winner team to be identified in each category.
- Scores for all teams will be published, identified only by inscription number. Experimental data will be released after the prize awarding (June 2025).
- Prizes will be awarded to the teams which achieve the highest score in each category (monotonic and cyclic tests). Two additional prizes will be awarded by the Organiser Committee taking into consideration the achieved score but also the composition of the contestant teams.
- The winning teams will be publicly announced on the event website and through the social media channels of GEOLAB on October 28, 2024. A representative from each winning team will be invited to participate in the 5th International Symposium on Frontiers in Offshore Geotechnics (ISFOG) in Nantes, France, (June 9-13, 2025), with covered registration fee and a travel stipend.
- The results of the BPC will be presented in a dedicated session at the conference by the organisers, followed by public recognition of the winners. The possibility of allowing the winning teams to present their approach at this prestigious conference is being explored.
- The organisers would like to thank ISFOG2025, the Association of Friend of the Institute of Geotechnics, and our industry sponsors (BAUER, IMPLENIA, KELLER and ZÜBLIN) for making this prize possible.
- The Organising Committee will prepare an international journal paper comparing and analysing the received predictions, with the goal of assessing the state of art and the state of practice.
- By participating the BPC, the team members authorise the use of the submitted data for that purpose. The participants BPC will not be invited as co-authors of this paper. The contestant teams are free to publish their predictions in separate papers using the experimental data provided, as long as the source is properly acknowledged.
Updated information on the BPC will be published in the event website until the announcement of the winner teams. If you would like to receive the updates via email, please sign in here for the BPC mailing list.
To participate in the BPC, teams must complete the provided templates: a spreadsheet for their predictions and a document outlining their methodology. Both documents should then be submitted by email to geolab_bpc@geotechnik.tu-darmstadt.de .
The deadline for submissions is September 30, 2024.
18 Mar 2024: First announcement and newsletter registration opening
6 May 2024: Publication of supporting documentation on the test setup and sand characterisation.
27 May 2024 : Publication of detailed participation rules and prediction sheet.
15 Sep 2024 : Closure of the submission period
30 Sep 2024: Closure of the submission period
28 Oct 2024: Announcement of the BPC winners
9-13 Jun 2025: Public recognition of BPC winners in ISFOG2025
The supporting documents for the BPC are listed below: | |||
File | Version | Format | Last updated on |
GEOLAB-BPC_01-TestSetupReport | v2 | 13.08.2024 | |
GEOLAB-BPC_02-TestSetupPictures | v1 | 07.05.2024 | |
GEOLAB-BPC_03-TestPitDrawing | v1 | 07.05.2024 | |
GEOLAB-BPC_04-LoadingSystemDrawing | v1 | 07.05.2024 | |
GEOLAB-BPC_05-PileCalibrationReport | v2 | 27.08.2024 | |
GEOLAB-BPC_06-CPTData | v1 | XLSX | 07.05.2024 |
GEOLAB-BPC_07-CPTLocationDrawing | v1 | 07.05.2024 | |
GEOLAB-BPC_08-SandCharacterisation | v7 | XLSX | 13.08.2024 |
GEOLAB-BPC_00-ParticipationRules | v2 | 15.10.2024 | |
GEOLAB-BPC_00-Prediction&ScoringSheet | v1 | XLSX | 28.05.2024 |
GEOLAB-BPC_00-PredictionMethod | v1 | DOCX | 28.05.2024 |
These documents can be downloaded here. Every effort has been made to prepare supporting documents as clearly and completely as possible. If you find anything unclear or require additional information, please contact the organisers by email. If any amendments to the supporting documents are necessary, they will be communicated on this webpage and via the BPC mailing list. |
Eighteen teams submitted predictions for the Monotonic Pile Test, with seven of those teams also submitting predictions for the Cyclic Pile Test.
In total, the contest had 59 participants from 15 countries, representing both industry and academia.
Monotonic Test Predictions | |||
Position | Team ID | Composition | Score (ouf of 100) |
1 | T10 | Researchers | 71.00 |
2 | T05 | Researchers | 69.20 |
3 | T17 | Researchers | 59.14 |
4 | T16 | Mixed | 54.38 |
5 | T09 | Researchers | 52.45 |
6 | T02 | Researchers | 47.49 |
7 | T14 | Researchers | 41.76 |
8 | T18 | Researchers | 37.86 |
9 | T11 | Researchers | 31.94 |
10 | T06 | Practice Engineers | 30.36 |
11 | T04 | Practice Engineers | 29.87 |
12 | T03 | Researchers | 28.93 |
13 | T08 | Mixed | 28.63 |
14 | T12 | Researchers | 24.91 |
15 | T13 | Researchers | 22.83 |
16 | T01 | Practice Engineers | 20.13 |
17 | T07 | Practice Engineers | 19.19 |
18 | T15 | Researchers | 14.92 |
Cyclic Test Predictions | |||
Position | Team ID | Composition | Score (out of 100) |
1 | T09 | Researchers | 78.65 |
2 | T10 | Researchers | 62.80 |
3 | T02 | Researchers | 59.83 |
4 | T05 | Researchers | 53.37 |
5 | T03 | Researchers | 49.44 |
6 | T17 | Researchers | 43.03 |
7 | T01 | Practice Engineers | 28.96 |
Combined scores (Montonic and Cyclic Tests) | |||
Position | Team ID | Composition | Score (out of 200) |
1 | T10 | Researchers | 133.81 |
2 | T09 | Researchers | 131.10 |
3 | T05 | Researchers | 122.57 |
4 | T02 | Researchers | 107.33 |
5 | T17 | Researchers | 102.17 |
6 | T03 | Researchers | 78.37 |
7 | T01 | Practice Engineers | 49.09 |
Best Monotonic Test Prediction and Best Overall Prediction: IC-MAGE team (T10) | ||
Pishun Tantivangphaisal | Imperial College London | United Kingdom |
Fabián Ortiz Wall | Imperial College London | United Kingdom |
David Taborda | Imperial College London | United Kingdom |
Best Cyclic Test Predicition and 2nd Best Overall Prediction: Offshore Geotechnical Group at Zhejiang University (T09) | ||
Lizhong Wang | Zhejiang University | China |
Xing Zha | Zhejiang University | China |
Yi Hong | Zhejiang University | China |
Yaru Zhang | Zhejiang University | China |
Junwei Liu | Qingdao University of Technology | China |
Zhen Guo | Zhejiang University | China |
Huan Wang | Norwegian Geotechnical Institute | China |
Rui Liu | Zhejiang University | China |
Xuetao Wang | Zhejiang University | China |
Lilin Wang | Zhejiang University | China |
3rd Best Overall Prediction: HCA Team (T05) | ||
José Duque | Universidad de la Costa | Colombia |
Carlos Lascarro | Universidad del Sinú | Colombia |
Merita Tafili | Ruhr-Universität Bochum | Germany |
David Mašín | Charles University | Czech Republic |
4th Best Overall Prediction: Centrif Eiffel (T02) | ||
Zheng Li | Gustave Eiffel University | France |
Matthieu Blanc | Gustave Eiffel University | France |
Luc Thorel | Gustave Eiffel University | France |
Organising Committee
Prof. Dr.-Ing. Hauke Zachert
Dr.-Ing. Jan Machaček
Dr.-Ing. Joaquín Liaudat
M.Sc. Simon Siegel
GEOLAB Project
The GEOLAB project is a collaborative initiative aimed at advancing geotechnical research and innovation through cutting-edge laboratory facilities and interdisciplinary cooperation. The GEOLAB project, funded by the European Union H2020 Research and Innovation Programme, spans four years from 2021 to 2025. The ten GEOLAB partners offer access to a wide range of experimental facilities, including:
- University of Maribor Large-scale Triaxial Apparatus
- Technical University of Darmstadt Pile Foundation Test Pit
- ETH Zürich Beam & Drum Centrifuge
- Norwegian Geotechnical Institute Geotest Sites
- Deltares GeoCentrifuge and GeoModel Container
- University of Cambridge Schofield Centre
- TU Delft Geotechnical Centrifuge and Liquefaction Tank
- CEDEX Track Box
- Gustave Eiffel University Geotechnical Centrifuge