PSI - Issue 38

Journal Logo Journal Logo Journal Logo l Journal Journal Logo

Available online at www.sciencedirect.com Structural Integrity Procedia 00 (2021) 1– ??

www.elsevier.com / locate / procedia

ScienceDirect Structural Integrity Procedia 00 (2021) 1– ?? tructural Integrity Proc dia 00 (2021) 1– ?? Structural Integrity Procedia 00 (2021) 1– ?? Structural Integrity Procedia 00 (2021) 1– ?? Structural Integrity Procedia 00 (2021) 1– ?? Procedia Structural Integrity 38 (2022) 497–506

www.elsevier.com / locate / procedia www.elsevier.com / locate / procedia www.elsevier.com / locate / procedia www.elsevier.com / locate / procedia www.elsevier.com / locate / procedia

Fatigue Design 2021, 9th Edition of the International Conference on Fatigue Design

2452-3216 © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Fatigue Design 2021 Organizers 10.1016/j.prostr.2022.03.050 ∗ Corresponding author: Email address: emilien.baroux@stellantis.com. (Emilien Baroux) ∗ Corresponding author: E ail address: emilien.baroux@stellantis.com. (E ilien Baroux) ∗ Corresponding author: Email address: emilien.baroux@stellantis.com. (Emilien Baroux) facing failure in service lies below the maximum acc ptable risk. This paradigm, opposed to maximalist design, aims to reduce costly system overdesigns (Svensson and Johannesson (2013)). Risk of failure of ∗ Corresponding author: Email address: emilien.baroux@stellantis.com. (Emilien Baroux) © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Fatigue Design 2021 Organizers Abstract In order to reliably design automotive structures, engineers need to determine and justify validation conditions and levels. These must stem from a thorough knowledge of structural damage induced by service loading conditions. From multi-input variable amplitude loading histories applied on a car’s wheel axles, we propose a multidimensional pseudo-damage description for the design of car chassis weak points. We present a multivariate description of client loading histories. We use it in a statistical analysis of a labelled measurement campaign to explain the heterogeneity of car driver profiles. Finally, we explore the question of complex load damage reconstruction using proving ground reference loads. Keywords: Automotive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multivariate analysis, Statistical learning, Loading reconstruction 1. Introduction Ensuring client safety, decreasing conception costs and reducing time-to-market are the three conflicting objectives at play when choosing and building processes for structural durability assessment. Requirements regarding the reliability of vehicle chassis parts seek to limit the probability (risk) of fatigue failure under regular (service) use cases over a given lifetime goal. Test conditions seek to validate such requirements when designing a new car model. They must be elaborated with respect to the uncertainties inherent to both fatigue failure, automotive industrial process and targeted customer uses (Johannesson (2014) Chap. 1). The choice of a design load must be accounted for. One method is to gather and compile real driving events, like in Standard Loading Histories (Berger et al. (2002) or Heuler and Kla¨tschke (2005)). When designing a part or organ, the chosen loading level must be proven representative of the whole population of in-service loads. Only thus can engineers control that the designed system’s probability of facing failure in service lies below the maximum acceptable risk. This paradigm, opposed to maximalist design, aims to reduce costly system overdesigns (Svensson and Johannesson (2013)). Risk of failure of Abstract In order to r liably design automotive structures, n ineer need to etermine and justify validati n conditions a d levels. These must stem from a thorough knowledge of structural damage induced by service loading conditi s. From multi-input variable amplitude loading historie applied on a car’s whe l axles, we propose a multidimensional pseudo-damage description for the design of car chassis weak points. We present a multivariate description of client loading histories. We use it in a stat stical analysis of a labelled measurement campaign to explain the heterogeneity of car driver profiles. Finally, we explore the question of complex load da age reconstruction using proving ground reference loads. Keywords: Automotive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multivariate analysis, Statistical learning, Loading reconstruction 1. Introduction Ensuring client safety, decreasing conception costs and reducing time-to-market are the three conflicting objectives at play when choosing and building processes for structural durability assessment. Requirements arding the reliability of vehicle chassis parts seek to limit the probability (risk) of fatigue failure under regular (service) use cases over a given lifetime goal. Test conditions seek to validate such requirements when designing a new car model. They must be elaborated with respect to the uncertainties inherent to both fatigue failure, automotive industrial process and targeted customer uses (Johannesson (2014) Chap. 1). Th choice of a design load must be accounted for. One method is to gather and compile real driving events, like in Standard Loading Histories (Berger et al. (2002) or Heuler and Kla¨tschke (2005)). When designing a part or organ, the chosen loading level must be proven representative of the whole population of in-service loads. Only thus can engineers control that the designed system’s probability of facing failure in service lies below the maximum acceptable risk. This paradigm, opposed to maximalist design, aims to reduce costly system overdesigns (Svensson and Johannesson (2013)). Risk of failure of Abstract In order to reliably design automotive structures, engineers need to determine and justify validation conditions and levels. These must stem from a thorough knowledge of structural damage induced by service loading conditions. From multi-input variable amplitude loading histories applied on a car’s wheel axles, we propose a multidimensional pseudo-damage description for the design of car chassis weak points. We present a multivariate description of client loading histories. We use it in a statistical analysis of a labelled measurement campaign to explain the heterogeneity of car driver profiles. Finally, we explore the question of complex load damage reconstruction using proving ground reference loads. Keywords: Aut motive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multiv iate analy is, Statistical learning, Loading reconstruction 1. Introducti n Ensuring client safety, decreasing conception costs and reducing time-to-market are the three conflicting objectives at play when choosing and building processes for structural durability assessment. Requirements regarding the reliability of vehicle chassis parts seek to limit the probability (risk) of fatigue failure under regular (service) use cases over a given lifetime goal. Test conditions seek to validate such requirements when designing a new car model. They must be elaborated with respect to the uncertainties inherent to both fatigue failure, automotive industrial process and targeted customer uses (Johannesson (2014) Chap. 1). The choice of a design load must be accounted for. One method is to gather and compile real driving events, like in Standard Loading Histories (Berger et al. (2002) or Heuler and Kla¨tschke (2005)). When designing a part or organ, the chosen loading level must be proven representative of the whole population of in-service loads. Only thus can engineers control that the designed system’s probability of facing failure in service lies below the maximum acceptable risk. This paradigm, opposed to maximalist design, aims to reduce costly system overdesigns (Svensson and Johannesson (2013)). Risk of failure of ∗ Corresponding author: Email address: emilien.baroux@st llantis.com. (Emilien Bar ux) roux a,b,c,d, ∗ , Benoit Delattre a , Andrei Constantinescu b , Patrick il c,d , I lt a a Stellantis, Technical Center Ve´lizy, Route de isy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de e´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de athe´ atiques d’ rsay, niversite´ Paris-Saclay, Bat 407, 91405 rsay Cedex 9, France d CELESTE tea Inria-Saclay, 1 Rue onore´ d’Estienne d’ rves, 91120 Palaiseau, France Abstract In order to reliably design automotive structures, engineers need to determine and justify validation conditions and levels. These must stem from a thorough knowledge of structural damage induced by service loading conditions. From multi-input variable amplitude loading histories applied on a car’s heel axles, e propose a ultidi ensional pseudo-damage description for the design of car chassis weak points. We present a ultivariate description of client loading histories. e use it in a statistical analysis of a labelled easure ent ca paign to explain the heterogeneity of car driver profiles. Finally, e explore the question of co lex load da age reconstruction using proving ground reference loads. Keywords: Automotive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multivariate analysis, Statistical learning, Loading reconstruction 1. Introduction nsuring client safety, decreasing conception costs and reducing ti e-to- arket are the three conflicting objectives at play hen cho sing and building processes for structural durability assess ent. i t regarding the reliability of vehicle chassis parts seek to li it the p obability (risk) of fatigue failure under regular (servic ) use cases over a given lifeti e goal. est conditi ns seek to validate such req ire ents when designing a n car odel. hey ust be elaborated ith resp ct to the uncertainties inh rent to both fatigue f ilure, auto otive industrial process and t rgeted custo r uses (Jo annesson (2014) hap. 1). he choic of a d sign load ust be account d for. ne ethod is to gather and co pile real dri ing events, lik in Standard oading istories ( erger et al. (2002) or euler and la¨tschke (2005)). hen designi a part or organ, the chosen loading level ust be proven representative of the hole population of in-service loads. nly thus can engineers control that the designed syste ’s probability f facing failure i i lies belo the axi u acceptable risk. his paradig , opposed to axi alist design, ai s to reduce costly syste overdesigns (Svensson and Johannesson (2013)). isk of failure of Analysis Of Real-Life Multi-Input Loading istories or he eliable esign Of Vehicle Chassis E ilien aroux a,b,c,d, ∗ , enoit elattre a , Andrei Constantinescu b , Patrick Pa phile c,d , Ida aoult a a Stellantis, Technical Center Ve´lizy, Route de Gisy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de Me´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de Mathe´matiques d’Orsay, Universite´ Paris-Saclay, Bat 407, 91405 Orsay Cedex 9, France d CELESTE team Inria-Saclay, 1 Rue Honore´ d’Estienne d’Orves, 91120 Palaiseau, France Abstract In order to r liably design automotive structures, engineers need to etermine and justify validation conditions a d levels. These must stem from a thorough knowledge of structural damage induced by service loading conditi s. From multi-input variable amplitude loading historie applied on a car’s whe l axles, we propose a multidimensional pseudo-damage description for the design of car chassis weak points. We present a multivariate description of client loading histories. We use it in a statistical analysis of a labelled measurement campaign to explain the heterogeneity of car driver profiles. Finally, we explore the question of complex load damage reconstruction using proving ground reference loads. Keywords: Automotive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multivariate analysis, Statistical learning, Loading reconstruction 1. Introduction Ensuring lient safety, decreasing conception c sts and reducing time-to-market are the three conflicting objectives at play when choosing and building processes for structural durability assessment. Requirements regarding the reliability of vehicle chassis p rts seek to limi the probability (risk) of fatigue failure under regular (service) use cases ov r a given li etime goal. Test conditions seek to validate such requirements when designing a new car model. They must be elaborat d with respect to the uncertainties inherent to both fatigue failure, automotive industrial process and targeted custo er uses (J ha nesson (2014) Chap. 1). Th choice of a design load must be accounted for. O e method is to gather and compile real driving events, like in Standard Loading Histories (Berger et al. (2002) or Heuler and Kla¨tschke (2005)). When designing a part or organ, the chosen loading level must be prove representative of the whole population of in-service loads. Only thus can engineers control that the designed system’s probability of facing failure in service lies below the maximum acceptable risk. This paradigm, opposed to maximalist design, aims to reduce costly system overdesigns (Svensson and Johannesson (2013)). Risk of failure of Fatigue Design 2021, 9th Edition of the International Conference on Fatigue Design Analysis Of al-Life Multi-Input Loading Histories For The Reliable Design Of Vehicle Chassis Emilien Baroux a,b,c,d, ∗ , Benoit Delattre a , Andrei Constantinescu b , Patrick Pamphile c,d , Ida Raoult a a Stellantis, T chnical Center Ve´lizy, Rou de Gisy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de Me´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de Mathe´matiques d’Orsay, Universit Paris-Saclay, Bat 407, 91405 Or ay Cedex 9, France d CELESTE team Inria-Saclay, 1 Rue Honore´ d’Estienne d’Orves, 91120 Palaiseau, France Abstract In order to reliably design automotive structures, engineers need to etermine and justify validation conditio and levels. These must stem from a thorough knowledge of structural damage induced by se vice loading conditi s. From multi-input variabl amplitude loading histories applied on a c r’s wheel axles, we propose a multidimensional pseudo- amage descript on for the design of car chassis we k oints. We present a multivariate description f clie t loading histories. We use i in a statistical analysis of a labelled measurement campaign to explai the heterogeneity of car driver profiles. Finall , we explore the question of complex load damage reconstruction using proving ground reference loads. Keywords: Automotive fatigue design, Multi-input variable amplitude loads, Multiaxial fatigue, Structural damage, Multivariate analysis, Statistical learni , Loading reconstruction 1. Introduction Ensur ng c ient safety, decr asing conce tion costs an reducing time-to-mark t are he thre conflicti g objectives at pl y when choosing and building process s for structural durability a sessment. R ir m ts arding the reliability of vehicle chassis parts s ek to limit the pr bability (risk) of fatigue fail re under regul r ( ervice) use ca es over a iven lifetime goal. Test co ition seek to validate such requiremen s when d signin a ew car model. They must b e aborated with r pect to t e unce tainties inherent to b th fatigue failure, automotive in ustria process a d targeted customer uses (Johannesson (2014) Chap. 1). Th choice of a d sign load must b account d for. One method is to gather and compile real driving events, lik in Standard Loadi Histories (Berger et al. (2002) or Heuler nd Kla¨tschke (2005)). When designing a part or organ, the chosen loading lev l must be proven representative f the whole population of -s oads. Only thus can engineers c trol that the designed system’s probability f Fatigue Design 2021, 9th Edition of the International Conference on Fatigue Design Fatigue Design 2021, 9th Edition of the International Conference on Fatigue Design Analysis Of Real-Life Multi-I put Loadi g Histories For The Reliable Design Of Vehicle Chassis Emilien Baroux a,b,c,d, ∗ , Benoit Delattre a , Andrei Constantinescu b , Patrick Pamphile c,d , Ida Raoult a a Stellantis, Te hnical Center Ve´lizy, Route de Gisy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de Me´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de Mathe´matiques d’Orsay, Universite´ Paris-Saclay, Bat 407, 91405 Orsay Cedex 9, France d CELESTE team I ri -Saclay, 1 Rue Honore´ d’Estienne d’Orves, 91120 Palaiseau, France ati e esign 2021, 9th Edition of the I ter ati al fere ce ati e esi l i f l- if lti-I ut i i t ri r eli l i f i l i ili Fatigue Design 2021, 9th Edition of the I ternational Co ference on Fatigue esign Analysis Of Real-Life Multi-Input Loading Histories For The Reliable Design Of Vehicle Chassis Emilien Baroux a,b,c,d, ∗ , Benoit Delattre a , Andrei Constantinescu b , Patrick Pamphile c,d , Ida Raoult a a Stellantis, Technical Center Ve´lizy, Route de Gisy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de Me´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de Mathe´matiques d’Orsay, Universite´ Paris-Saclay, Bat 407, 91405 Orsay Cedex 9, France d CELESTE team Inria-Saclay, 1 Rue Honore´ d’Estienne d’Orves, 91120 Palaiseau, France Analysis Of Real-Life Multi-Input Loading Histories For The Reliable Design Of Vehicle Chassis Emilien Baroux a,b,c,d, ∗ , Benoit Delattre a , Andrei Constantinescu b , Patrick Pamphile c,d , Ida Raoult a a Stellantis, Technical Center Ve´lizy, Route de Gisy, 78140 Ve´lizy-Villacoublay, France b Laboratoire de Me´canique des Solides, Institut Polytechnique de Paris, Ecole Polytechnique, 91120 Palaiseau, France c Laboratoire de Mathe´matiques d’Orsay, Universite´ Paris-Saclay, Bat 407, 91405 Orsay Cedex 9, France d CELESTE team Inria-Saclay, 1 Rue Honore´ d’Estienne d’Orves, 91120 Palaiseau, France ∗ Corresponding author: Email address: emilien.baroux@stellantis.com. (Emilien Baroux)