Pub­li­ca­tions

Cur­rent publications

RESEARCH ARTI­CLE:
MUL­TI­SPEC­TRAL EVALUATION

Ana­lyz­ing mul­ti­spec­tral emis­sion and syn­chro­tron data to eval­u­ate the qual­i­ty of laser welds on cop­per
(Jan Brüggen­jür­gen, Christoph Spurk, Marc Hum­mel, Christoph Franz, Andrè Häusler, Alexan­der Olowin­sky, Felix Beck­mann, Julian Moos­mann; Ana­lyz­ing mul­ti­spec­tral emis­sion and syn­chro­tron data to eval­u­ate the qual­i­ty of laser welds on cop­per. J. Laser Appl. 1 August 2024; 36 (3): 032032.) 

Descrip­tion
The val­i­da­tion of laser weld­ing of metal­lic mate­ri­als is a chal­leng­ing task due to the high­ly dynam­ic process­es and the lim­it­ed acces­si­bil­i­ty of the weld seam. Mea­sur­ing the process emis­sions and the laser beam being processed is one way of record­ing high­ly dynam­ic process phe­nom­e­na.
How­ev­er, these record­ings are always made via the sur­face of the weld seam, mean­ing that phe­nom­e­na inside the seam are only implic­it­ly rec­og­niz­able and require fur­ther processing.

Part­ner
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT
> RWTH Aachen Uni­ver­si­ty
> Insti­tute of Mate­ri­als Physics

Learn more
Learn more
RESEARCH ARTI­CLE:
MUL­TI­SPEC­TRAL MONITORING

Mul­ti-spec­tral Mon­i­tor­ing Data in Cor­re­la­tion with X‑ray Videog­ra­phy dur­ing Laser Weld­ing of Hair­pins (Christoph Franz, Sören Hol­latz, Oliv­er Burch­wald, Christoph Spurk, Marc Hum­mel, Alexan­der Olowin­sky, Eve­line Rein­heimer, Chris­t­ian Hagen­locher, Felix Beck­mann, Julian Moos­mann; 13th CIRP Con­fer­ence on Pho­ton­ic Tech­nolo­gies [LANE 2024])

Descrip­tion
In today’s pro­duc­tion of high-per­for­mance elec­tric engines, the hair­pin tech­nol­o­gy is used to increase the effi­cien­cy. Instead
of a sta­tor made of wind­ed round wire, thick­er cop­per pins are assem­bled and weld­ed. Typ­i­cal weld fail­ures such as spat­ters,
pores or insuf­fi­cient con­nec­tion occur due to sur­face con­t­a­m­i­na­tion, insuf­fi­cient clamp­ing, posi­tion­ing or pre­vi­ous cut­ting
process­es. For pro­duc­tion facil­i­ties, it is not suf­fi­cient to iden­ti­fy defec­tive welds; a clas­si­fi­ca­tion is also required in order to
deter­mine the caus­es of the fault and rec­ti­fy them as soon as pos­si­ble. The capa­bil­i­ty of a mul­ti-spec­tral mon­i­tor­ing sys­tem is
eval­u­at­ed in this study with the help of in-situ X‑ray videog­ra­phy. The data show a cor­re­la­tion with the sta­bil­i­ty of the vapour
cap­il­lary, the weld­ing posi­tion and spat­ter formation.

Part­ners
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT
> RWTH Aachen Uni­ver­si­ty
> Uni­ver­si­ty of Stuttgart, Insti­tut für Strahlw­erkzeuge (IFSW)
> Insti­tute of Mate­ri­als Physics

RESEARCH ARTI­CLE:
INFLU­ENCE ON PROCESS RADIATION

Influ­ence of full pen­e­tra­tion weld­ing on the mul­ti­spec­tral sig­nal of the
process radi­a­tion dur­ing laser beam weld­ing of austenitic stain­less steel (Patrick Mey­er, Chris­t­ian Schumann,Dario Spatzek, André Häusler; 13th CIRP Con­fer­ence on Pho­ton­ic Tech­nolo­gies [LANE 2024])

Descrip­tion
A 100% visu­al inspec­tion of the welds car­ried out is not always pos­si­ble for eco­nom­ic rea­sons. Dif­fer­ent meth­ods of process mon­i­tor­ing are there­fore used for non-destruc­tive test­ing of process and weld seam qual­i­ty. In the fol­low­ing, the record­ed mul­ti­spec­tral sig­nals of the reflec­tive process radi­a­tion are eval­u­at­ed using descrip­tive sta­tis­tics and char­ac­ter­is­tics of a full pen­e­tra­tion weld dur­ing laser beam weld­ing of austenitic stain­less steel are worked out. The aim is to detect and local­ize unin­ten­tion­al full pen­e­tra­tion weld­ing at an ear­ly stage and thus avoid rejects. For this pur­pose, adapt­ed sam­ple geome­tries with vary­ing mate­r­i­al thick­ness­es are pre­pared for the test series. As a result, the need to vary the weld­ing process para­me­ters to achieve the dif­fer­ent weld­ing states (full and par­tial pen­e­tra­tion weld) is avoid­ed. By eval­u­at­ing the mea­sured val­ues of the wave­length-selec­tive sig­nals, a sig­nif­i­cant influ­ence on the sig­nal inten­si­ty and the sig­nal fre­quen­cy could be confirmed.

Part­ners
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT

Learn more
Learn more
RESEARCH ARTI­CLE BY PROJECT PART­NERS:
WELD DEFECT DETEC­TION WITH 4D.TWO

Weld Defect Detec­tion in Laser Beam Weld­ing Using Mul­ti­spec­tral Emis­sion Sen­sor Fea­tures and Machine Learn­ing (Ame­na Dar­wish, Man­fred Pers­son, Ste­fan Eric­son, Rohol­lah Ghase­mi, Kent Salomonsson)

Descrip­tion
Laser beam weld­ing is a high­ly com­plex process in which even the slight­est devi­a­tions can lead to hard-to-detect defects such as poros­i­ty. In this arti­cle, researchers from the Uni­ver­si­ty of Skövde present an inno­v­a­tive, data-dri­ven frame­work for detect­ing weld­ing defects using mul­ti­spec­tral emis­sion sen­sors and machine learn­ing. By com­bin­ing super­vised and unsu­per­vised learn­ing meth­ods, the sys­tem suc­cess­ful­ly extracts rel­e­vant fea­tures from 4D pho­to­di­ode sig­nals and cor­re­lates them with weld­ing defects – both for post-process analy­sis and real-time mon­i­tor­ing. The study demon­strates how AI-based eval­u­a­tion of sen­sor sig­nals can rev­o­lu­tion­ize qual­i­ty assur­ance in indus­tri­al manufacturing.

RESEARCH ARTI­CLE BY PROJECT PART­NERS:
MON­I­TOR­ING DUR­ING LASER BEAM WELDING

Sto­chas­tic analy­sis of mul­ti­spec­tral data dur­ing laser beam weld­ing of stain­less steel with dif­fer­ent gap sizes (Patrick Mey­er, André Häusler, Alexan­der Olowin­sky; IOP Con­fer­ence Series: Mate­ri­als Sci­ence and Engi­neer­ing, Vol­ume 1332, 20th Nordic Laser Mate­ri­als Pro­cess­ing Con­fer­ence 26/08/2025 — 28/08/2025 Kon­gens Lyn­g­by, Denmark)

Descrip­tion
Qual­i­ty assur­ance in laser beam weld­ing demands high lev­els of pre­ci­sion and respon­sive­ness. This paper intro­duces an inno­v­a­tive approach to process mon­i­tor­ing using pho­to­di­ode sen­sors that cap­ture reflect­ed laser radi­a­tion. The sen­sor data pro­vides valu­able insights into the weld­ing process and, when com­bined with machine learn­ing, enables reli­able detec­tion of process devi­a­tions. The goal is to mon­i­tor weld qual­i­ty in real time and iden­ti­fy poten­tial defects at an ear­ly stage – a cru­cial step toward auto­mat­ed and intel­li­gent manufacturing.

Learn more
Learn more
THE­SIS BY PROJECT PART­NERS:
MOD­I­FI­CA­TION OF MATE­R­I­AL PROPERTIES

Pilot Study for Data-dri­ven Mod­i­fi­ca­tion of Face Spe­cif­ic Mate­r­i­al Prop­er­ties (Man­fred Persson)

Descrip­tion
Pre­cise mon­i­tor­ing of laser process­es is cru­cial for ensur­ing the qual­i­ty of mod­ern man­u­fac­tur­ing. This study presents an inno­v­a­tive approach for real-time eval­u­a­tion of weld qual­i­ty in alu­minum bus­bar laser weld­ing. By uti­liz­ing mul­ti­spec­tral pho­to­di­ode sen­sors and deep learn­ing algo­rithms, rel­e­vant process data can be cap­tured and ana­lyzed instant­ly. The sys­tem detects qual­i­ty devi­a­tions ear­ly and enables auto­mat­ed process opti­miza­tion – a key step toward Indus­try 4.0 and intel­li­gent manufacturing.

THE­SIS BY PROJECT PART­NERS:
REAL-TIME QUAL­I­TY PREDICTION

Real-time qual­i­ty pre­dic­tion in bus­bar laser weld­ing: A mul­ti-spec­tral pho­to­di­ode and Con­vL­STM frame­work (Dan Lönn)

Descrip­tion
Pre­cise mon­i­tor­ing of laser process­es is essen­tial for the qual­i­ty of safe­ty-crit­i­cal com­po­nents, such as those used in the auto­mo­tive indus­try. This mas­ter’s the­sis presents an inno­v­a­tive sys­tem for real-time eval­u­a­tion of weld qual­i­ty in alu­minum bus­bar laser weld­ing. It com­bines mul­ti­spec­tral pho­to­di­ode sens­ing with a deep learn­ing mod­el (Con­vL­STM) to ana­lyze back-reflec­tions and spec­tral sig­nals dur­ing the weld­ing process. Light in the range of 400 nm to 700 nm pro­vides valu­able insights into mate­r­i­al changes and poten­tial defects. The result is an intel­li­gent, data-dri­ven solu­tion for ear­ly detec­tion of weld­ing flaws and opti­miza­tion of indus­tri­al man­u­fac­tur­ing processes.

Learn more
RESEARCH ARTI­CLE: MUL­TI­SPEC­TRAL EVALUATION
Ana­lyz­ing mul­ti­spec­tral emis­sion and syn­chro­tron data to eval­u­ate the qual­i­ty of laser welds on cop­per (Jan Brüggen­jür­gen, Christoph Spurk, Marc Hum­mel, Christoph Franz, Andrè Häusler, Alexan­der Olowin­sky, Felix Beck­mann, Julian Moos­mann; Ana­lyz­ing mul­ti­spec­tral emis­sion and syn­chro­tron data to eval­u­ate the qual­i­ty of laser welds on cop­per. J. Laser Appl. 1 August 2024; 36 (3): 032032.)
 

Descrip­tion
The val­i­da­tion of laser weld­ing of metal­lic mate­ri­als is a chal­leng­ing task due to the high­ly dynam­ic process­es and the lim­it­ed acces­si­bil­i­ty of the weld seam.
Mea­sur­ing the process emis­sions and the laser beam being processed is one way of record­ing high­ly dynam­ic process phe­nom­e­na.
How­ev­er, these record­ings are always made via the sur­face of the weld seam, mean­ing that phe­nom­e­na inside the seam are only implic­it­ly rec­og­niz­able and require fur­ther processing.

Part­ner
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT
> RWTH Aachen Uni­ver­si­ty
> Insti­tute of Mate­ri­als Physics

Learn more
RESEARCH ARTI­CLE:
MUL­TI­SPEC­TRAL MONITORING

Mul­ti-spec­tral Mon­i­tor­ing Data in Cor­re­la­tion with X‑ray Videog­ra­phy dur­ing Laser Weld­ing of Hair­pins (Christoph Franz, Sören Hol­latz, Oliv­er Burch­wald, Christoph Spurk, Marc Hum­mel, Alexan­der Olowin­sky, Eve­line Rein­heimer, Chris­t­ian Hagen­locher, Felix Beck­mann, Julian Moos­mann; 13th CIRP Con­fer­ence on Pho­ton­ic Tech­nolo­gies [LANE 2024])

Descrip­tion
In today’s pro­duc­tion of high-per­for­mance elec­tric engines, the hair­pin tech­nol­o­gy is used to increase the effi­cien­cy. Instead of a sta­tor made of wind­ed round wire, thick­er cop­per pins are assem­bled and weld­ed. Typ­i­cal weld fail­ures such as spat­ters, pores or insuf­fi­cient con­nec­tion occur due to sur­face con­t­a­m­i­na­tion, insuf­fi­cient clamp­ing, posi­tion­ing or pre­vi­ous cut­ting process­es. For pro­duc­tion facil­i­ties, it is not suf­fi­cient to iden­ti­fy defec­tive welds; a clas­si­fi­ca­tion is also required in order to deter­mine the caus­es of the fault and rec­ti­fy them as soon as pos­si­ble. The capa­bil­i­ty of a mul­ti-spec­tral mon­i­tor­ing sys­tem is eval­u­at­ed in this study with the help of in-situ X‑ray videog­ra­phy. The data show a cor­re­la­tion with the sta­bil­i­ty of the vapour cap­il­lary, the weld­ing posi­tion and spat­ter formation.

Part­ners
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT
> RWTH Aachen Uni­ver­si­ty
> Uni­ver­si­ty of Stuttgart, Insti­tut für Strahlw­erkzeuge (IFSW)
> Insti­tute of Mate­ri­als Physics

 

Learn more
RESEARCH ARTI­CLE:
INFLU­ENCE ON PROCESS RADIATION

Influ­ence of full pen­e­tra­tion weld­ing on the mul­ti­spec­tral sig­nal of the
process radi­a­tion dur­ing laser beam weld­ing of austenitic stain­less steel (Patrick Mey­er, Chris­t­ian Schumann,Dario Spatzek, André Häusler; 13th CIRP Con­fer­ence on Pho­ton­ic Tech­nolo­gies [LANE 2024])

Descrip­tion
A 100% visu­al inspec­tion of the welds car­ried out is not always pos­si­ble for eco­nom­ic rea­sons. Dif­fer­ent meth­ods of process mon­i­tor­ing are there­fore used for non-destruc­tive test­ing of process and weld seam qual­i­ty. In the fol­low­ing, the record­ed mul­ti­spec­tral sig­nals of the reflec­tive process radi­a­tion are eval­u­at­ed using descrip­tive sta­tis­tics and char­ac­ter­is­tics of a full pen­e­tra­tion weld dur­ing laser beam weld­ing of austenitic stain­less steel are worked out. The aim is to detect and local­ize unin­ten­tion­al full pen­e­tra­tion weld­ing at an ear­ly stage and thus avoid rejects. For this pur­pose, adapt­ed sam­ple geome­tries with vary­ing mate­r­i­al thick­ness­es are pre­pared for the test series. As a result, the need to vary the weld­ing process para­me­ters to achieve the dif­fer­ent weld­ing states (full and par­tial pen­e­tra­tion weld) is avoid­ed. By eval­u­at­ing the mea­sured val­ues of the wave­length-selec­tive sig­nals, a sig­nif­i­cant influ­ence on the sig­nal inten­si­ty and the sig­nal fre­quen­cy could be confirmed.

Part­ners
> Fraun­hofer Insti­tute for Laser Tech­nol­o­gy ILT

Learn more
RESEARCH ARTI­CLE BY PROJECT PART­NERS:
WELD DEFECT DETEC­TION WITH 4D.TWO

Weld Defect Detec­tion in Laser Beam Weld­ing Using Mul­ti­spec­tral Emis­sion Sen­sor Fea­tures and Machine Learn­ing (Ame­na Dar­wish, Man­fred Pers­son, Ste­fan Eric­son, Rohol­lah Ghase­mi, Kent Salomonsson)

Descrip­tion
Laser beam weld­ing is a high­ly com­plex process in which even the slight­est devi­a­tions can lead to hard-to-detect defects such as poros­i­ty. In this arti­cle, researchers from the Uni­ver­si­ty of Skövde present an inno­v­a­tive, data-dri­ven frame­work for detect­ing weld­ing defects using mul­ti­spec­tral emis­sion sen­sors and machine learn­ing. By com­bin­ing super­vised and unsu­per­vised learn­ing meth­ods, the sys­tem suc­cess­ful­ly extracts rel­e­vant fea­tures from 4D pho­to­di­ode sig­nals and cor­re­lates them with weld­ing defects – both for post-process analy­sis and real-time mon­i­tor­ing. The study demon­strates how AI-based eval­u­a­tion of sen­sor sig­nals can rev­o­lu­tion­ize qual­i­ty assur­ance in indus­tri­al manufacturing.

Learn more
RESEARCH ARTI­CLE BY PROJECT PART­NERS:
MON­I­TOR­ING DUR­ING LASER BEAM WELDING

Sto­chas­tic analy­sis of mul­ti­spec­tral data dur­ing laser beam weld­ing of stain­less steel with dif­fer­ent gap sizes (Patrick Mey­er, André Häusler, Alexan­der Olowin­sky; IOP Con­fer­ence Series: Mate­ri­als Sci­ence and Engi­neer­ing, Vol­ume 1332, 20th Nordic Laser Mate­ri­als Pro­cess­ing Con­fer­ence 26/08/2025 — 28/08/2025 Kon­gens Lyn­g­by, Denmark)

Descrip­tion
Qual­i­ty assur­ance in laser beam weld­ing demands high lev­els of pre­ci­sion and respon­sive­ness. This paper intro­duces an inno­v­a­tive approach to process mon­i­tor­ing using pho­to­di­ode sen­sors that cap­ture reflect­ed laser radi­a­tion. The sen­sor data pro­vides valu­able insights into the weld­ing process and, when com­bined with machine learn­ing, enables reli­able detec­tion of process devi­a­tions. The goal is to mon­i­tor weld qual­i­ty in real time and iden­ti­fy poten­tial defects at an ear­ly stage – a cru­cial step toward auto­mat­ed and intel­li­gent manufacturing.

Learn more
THE­SIS BY PROJECT PART­NERS:
MOD­I­FI­CA­TION OF MATE­R­I­AL PROPERTIES

Pilot Study for Data-dri­ven Mod­i­fi­ca­tion of Face Spe­cif­ic Mate­r­i­al Prop­er­ties (Man­fred Persson)

Descrip­tion
Pre­cise mon­i­tor­ing of laser process­es is cru­cial for ensur­ing the qual­i­ty of mod­ern man­u­fac­tur­ing. This study presents an inno­v­a­tive approach for real-time eval­u­a­tion of weld qual­i­ty in alu­minum bus­bar laser weld­ing. By uti­liz­ing mul­ti­spec­tral pho­to­di­ode sen­sors and deep learn­ing algo­rithms, rel­e­vant process data can be cap­tured and ana­lyzed instant­ly. The sys­tem detects qual­i­ty devi­a­tions ear­ly and enables auto­mat­ed process opti­miza­tion – a key step toward Indus­try 4.0 and intel­li­gent manufacturing.

Learn more
THE­SIS BY PROJECT PART­NERS:
REAL-TIME QUAL­I­TY PREDICTION

Real-time qual­i­ty pre­dic­tion in bus­bar laser weld­ing: A mul­ti-spec­tral pho­to­di­ode and Con­vL­STM frame­work (Dan Lönn)

Descrip­tion
Pre­cise mon­i­tor­ing of laser process­es is essen­tial for the qual­i­ty of safe­ty-crit­i­cal com­po­nents, such as those used in the auto­mo­tive indus­try. This mas­ter’s the­sis presents an inno­v­a­tive sys­tem for real-time eval­u­a­tion of weld qual­i­ty in alu­minum bus­bar laser weld­ing. It com­bines mul­ti­spec­tral pho­to­di­ode sens­ing with a deep learn­ing mod­el (Con­vL­STM) to ana­lyze back-reflec­tions and spec­tral sig­nals dur­ing the weld­ing process. Light in the range of 400 nm to 700 nm pro­vides valu­able insights into mate­r­i­al changes and poten­tial defects. The result is an intel­li­gent, data-dri­ven solu­tion for ear­ly detec­tion of weld­ing flaws and opti­miza­tion of indus­tri­al man­u­fac­tur­ing processes.

Learn more

Con­tact us!

If you would like to find out more about our prod­ucts and ser­vices or request a cus­tomized quote, please con­tact us. You can reach us by phone, e‑mail or via our con­tact form. We look for­ward to your inquiry!

Con­tact
Sched­ule con­sul­ta­tion appointment

Con­tact us!

If you would like to find out more about our prod­ucts and ser­vices or request a cus­tomized quote, please con­tact us. You can reach us by phone, e‑mail or via our con­tact form. We look for­ward to your inquiry!

Con­tact
Sched­ule con­sul­ta­tion appointment
Scroll to Top