Polymer-based Capacitive Micromachined Transducer-Enabled Inline Monitoring of Ultrasonic Welding in Thermoplastic Carbon Fiber Composites

Andrea Cossettini; Dominik Goerick; Edmond Cretu; Heinz Voggenreiter; Jinhao Lu; Jonas Welsch; Luca Benini; Martin Angerer; Michael Kupke; Robert Rohling

arxiv: 2606.12695 · v1 · pith:AUVJJ4YSnew · submitted 2026-06-10 · 📡 eess.SY · cs.SY

Polymer-based Capacitive Micromachined Transducer-Enabled Inline Monitoring of Ultrasonic Welding in Thermoplastic Carbon Fiber Composites

Jonas Welsch , Dominik Goerick , Martin Angerer , Jinhao Lu , Sergei Vostrikov , Michael Kupke , Heinz Voggenreiter , Andrea Cossettini

show 3 more authors

Luca Benini Edmond Cretu Robert Rohling

This is my paper

Pith reviewed 2026-06-27 08:11 UTC · model grok-4.3

classification 📡 eess.SY cs.SY

keywords ultrasonic weldingthermoplastic compositespolyCMUTinline monitoringnon-destructive testingdefect detectioncarbon fiber laminatesreal-time sensing

0 comments

The pith

Polymer-based CMUT sensors detect every induced defect in real-time ultrasonic welding of thermoplastic composites.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents a compact wireless ultrasonic system that integrates custom polymer-based capacitive micromachined transducers into an industrial welding setup for continuous inline monitoring. It reports that an eight-element array at roughly 3.6 MHz captures process-synchronous echo data showing consistent depth shifts exactly where defects were placed. Across 21 welds these shifts identify all defects with no misses and only limited false positives, matching X-ray computed tomography images. The work targets the quality-assurance gap in high-throughput production of lightweight recyclable composite structures. The sensors operate at low power inside the high-interference welding environment without requiring post-process inspection.

Core claim

An eight-element linear polyCMUT array at approximately 3.6 MHz, packaged and integrated with the ultra-low-power WULPUS platform, performs inline ultrasonic measurements during continuous welding of carbon-fiber laminates containing intentional defects; the resulting depth-of-echo shifts occur at every defect location, agree with X-ray CT ground truth, and allow detection of all defects across 21 welds with no false negatives and limited false positives.

What carries the argument

Eight-element linear polyCMUT array at center frequency of approximately 3.6 MHz integrated with the WULPUS platform for process-synchronous ultrasonic data acquisition inside the welding environment.

If this is right

Real-time defect detection becomes feasible inside the welding process itself.
Quality assurance can shift from post-weld inspection to in-process monitoring.
Polymer-based transducers enable low-cost, scalable sensing compatible with existing manufacturing lines.
The approach supports high-throughput production of recyclable thermoplastic composite structures.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same sensor integration could be adapted to monitor other continuous joining processes such as resistance welding or induction welding.
Wireless low-power operation opens the possibility of embedding multiple arrays along longer weld paths without cabling constraints.
If depth-of-echo shifts prove repeatable across material batches, the method might allow closed-loop control of welding parameters.
Extending the array length or adding beam-forming could increase spatial resolution for smaller defects.

Load-bearing premise

The observed echo shifts are produced specifically by the introduced defects rather than ordinary process variations or sensor placement, and the polyCMUT integration stays stable without unaccounted interference throughout the weld.

What would settle it

A weld containing a known defect that produces no measurable echo shift, or a defect-free weld that triggers repeated false-positive detections under the same sensor placement.

Figures

Figures reproduced from arXiv: 2606.12695 by Andrea Cossettini, Dominik Goerick, Edmond Cretu, Heinz Voggenreiter, Jinhao Lu, Jonas Welsch, Luca Benini, Martin Angerer, Michael Kupke, Robert Rohling, Sergei Vostrikov.

**Figure 8.** Figure 8: shows the measurement setup, the transmit sensitivity across 16 elements, and an example of the frequency- and angle-dependent response. The latter allows to assess the generated acoustic beam in 2D. Center frequency fC, fractional bandwidth (FBW), and peak sensitivity SC are listed in [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗

**Figure 9.** Figure 9: Evaluation of signal time of flight depending on material to validate welding seam location. a) signal with PEEK coupling block, b) signal with PEEK and first CFRP plate, c) signal of PEEK block and both coupled CFRP plates. A single plate shifted the echo by 1.125 µs at a sampling rate of 8 MSps, corresponding to a speed of sound of 2986 m/s at a plate thickness of 1.68 mm. The second plate had an average… view at source ↗

**Figure 10.** Figure 10: Top: Welding signals plotted over position with color coded intensity over depth of echo (similar to a B-Scan). Section a) is not usable due to unreliable surface contact, Section b) is a measurement during normal welding operation and section c) is the measurement while the sonotrode was already shut off. Bottom: High resolution CT scan of the same specimen with porosities visible after the second defect… view at source ↗

**Figure 12.** Figure 12: Section (b) - measurement during the running welding process with the four false positives clearly visible at 75 and 115 mm. All 20 measurements captured the intentionally induced defect. 3.6. System Evaluation & Future Work In the evaluation of non-destructive testing (NDT) systems, two types of misidentifications are commonly considered: false negatives, where the system fails to detect an existing defe… view at source ↗

read the original abstract

Thermoplastic composite structures enable lightweight, recyclable, and high-throughput aerospace manufacturing, but reliable quality assurance of advanced joining processes remains a key challenge. This work presents a compact, low-cost, and wireless ultrasonic non-destructive testing system for real-time, inline monitoring of continuous ultrasonic welding of thermoplastic carbon fiber composites. The system integrates custom-fabricated polymer-based capacitive micromachined ultrasonic transducers (polyCMUTs) with the ultra-low-power WULPUS platform, enabling operation in the harsh, high-interference welding environment. An eight-element linear polyCMUT array operating at a center frequency of approximately 3.6 MHz is designed, fabricated, packaged, and integrated into an industrial welding setup. Inline measurements are performed during welding of carbon fiber laminates with intentionally introduced defects. Process-synchronous ultrasonic data reveal consistent depth-of-echo shifts at defect locations, in strong agreement with X-ray computed tomography ground truth. Across 21 welds, all induced defects are detected without false negatives and with limited false positives. The results demonstrate that polymer-based CMUT technology enables robust, scalable, and manufacturing-compatible ultrasonic sensing, representing a decisive step toward intelligent process monitoring and quality assurance for next-generation thermoplastic composite welding.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

PolyCMUT integration gives a working inline monitor for ultrasonic welding with X-ray agreement on 21 samples, but the echo-shift attribution to defects needs baseline controls to hold up.

read the letter

The main thing here is a hardware demo that puts polymer CMUTs on the WULPUS platform into an actual continuous ultrasonic welding line for thermoplastic carbon fiber laminates. They report detecting all induced defects across 21 welds with no false negatives and limited false positives, and the depth-of-echo shifts line up with post-weld X-ray CT.

What stands out as new is the specific packaging and integration of an 8-element 3.6 MHz polyCMUT array into the welding setup so it runs process-synchronously in the high-interference environment. The abstract shows they fabricated, packaged, and operated the transducers without obvious failure during the runs.

The experimental comparison to independent CT ground truth is the strongest part. That gives the detection numbers some external anchor rather than relying on simulation or fitting.

The soft spot is exactly the one the stress-test note flags. Ultrasonic welding involves fast temperature, pressure, and flow changes that shift acoustic impedance on their own. The abstract ties the observed shifts to defect locations but does not describe defect-free control welds, quantitative variability measures in clean segments, or any statistical isolation of defect effects from normal process drift or sensor placement. If those baselines are missing or weak in the full text, the zero-false-negative claim rests on an untested assumption.

This paper is for engineers in composite process monitoring and NDT who need practical sensor examples rather than new transducer physics. Readers working on aerospace or recyclable thermoplastic joining would get the most from the integration details and the reported numbers.

It deserves a serious referee because the application is timely and the central claim is falsifiable with the right controls. Send it out, but ask specifically for the baseline data and variability analysis in revision.

Referee Report

1 major / 2 minor

Summary. The manuscript presents a compact, low-cost wireless ultrasonic NDT system using an eight-element polymer-based capacitive micromachined ultrasonic transducer (polyCMUT) array at ~3.6 MHz integrated with the WULPUS platform for real-time inline monitoring during continuous ultrasonic welding of thermoplastic carbon fiber composites. Inline process-synchronous measurements on 21 welds containing intentionally introduced defects show consistent depth-of-echo shifts at defect locations that agree with X-ray CT ground truth, achieving detection of all defects with zero false negatives and limited false positives. The work claims this demonstrates robust, scalable, manufacturing-compatible sensing for quality assurance in aerospace composite joining.

Significance. If the central experimental claims hold after addressing controls, the result would be significant for enabling intelligent, real-time process monitoring in high-throughput thermoplastic composite manufacturing. Strengths include the use of custom polyCMUT technology suited to harsh environments, wireless low-power operation, and direct comparison to independent X-ray CT ground truth across multiple welds.

major comments (1)

[Results (inline measurements and detection performance)] The central claim that observed depth-of-echo shifts are caused specifically by the introduced defects (rather than normal process variations in temperature, pressure, or polymer flow) is load-bearing but insufficiently supported. The results section provides no quantitative baseline comparison of echo variability in defect-free weld segments, no control welds without defects, and no statistical test isolating defect effects from sensor placement drift or process fluctuations. Without these, the reported zero false negatives and limited false positives cannot be confidently attributed to defect detection.

minor comments (2)

[Abstract and Results] The abstract and results should report the exact number and nature of false positives, along with any error bars or variability metrics on the echo-shift measurements.
[Methods] Methods details on polyCMUT integration stability, exclusion criteria for welds, and exact synchronization of ultrasonic data with the welding process are needed for reproducibility.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback and positive assessment of the work's significance. We address the single major comment below and will revise the manuscript accordingly to strengthen the attribution of the observed signals.

read point-by-point responses

Referee: The central claim that observed depth-of-echo shifts are caused specifically by the introduced defects (rather than normal process variations in temperature, pressure, or polymer flow) is load-bearing but insufficiently supported. The results section provides no quantitative baseline comparison of echo variability in defect-free weld segments, no control welds without defects, and no statistical test isolating defect effects from sensor placement drift or process fluctuations. Without these, the reported zero false negatives and limited false positives cannot be confidently attributed to defect detection.

Authors: We acknowledge the referee's concern that the results section would benefit from more explicit quantitative support for attributing the depth-of-echo shifts specifically to the defects. The manuscript already shows that these shifts occur consistently and exclusively at the positions of the intentionally introduced defects, with precise spatial agreement to the independent X-ray CT ground truth across all 21 welds and zero false negatives. The process-synchronous nature of the measurements further helps account for general process fluctuations. Nevertheless, we agree that adding a quantitative baseline of echo variability in defect-free segments, along with a statistical comparison, would strengthen the claim. In the revised manuscript we will include such an analysis drawn from the existing weld data (standard deviation of echo depths in defect-free regions versus defect locations) and add an appropriate statistical test. We will also explicitly note the lack of separate defect-free control welds as a limitation while explaining how the multi-weld consistency and CT validation address process variations. These changes will appear in the results and discussion sections. revision: yes

Circularity Check

0 steps flagged

No circularity: experimental validation against independent CT ground truth

full rationale

The paper describes fabrication, integration, and experimental testing of polyCMUT sensors during ultrasonic welding, with defect detection performance evaluated by direct comparison to X-ray CT measurements on the same welds. No mathematical derivations, equations, parameter fitting, or predictions are presented. No self-citations are invoked as load-bearing for any uniqueness theorem or ansatz. The central claims rest on empirical agreement with an external, independent measurement modality (CT), satisfying the criteria for a self-contained experimental result with no reduction to inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work is an experimental engineering demonstration with no free parameters fitted to data, no new postulated entities, and reliance on standard domain assumptions about ultrasonic propagation and defect echo formation.

axioms (1)

domain assumption Ultrasonic echo depth shifts reliably indicate weld defects in carbon fiber laminates under the tested conditions
Invoked to link observed signals to defect locations and to claim agreement with X-ray CT.

pith-pipeline@v0.9.1-grok · 5790 in / 1179 out tokens · 18239 ms · 2026-06-27T08:11:42.339592+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

Global Market Forecast 2025,

Introduction The aerospace industry is projected to experience unprecedented growth, with Airbus and Boeing each forecasting the delivery of over 43,000 new aircraft within the next two decades. [1,2] To limit environmental impact and operating costs, reducing fuel consumption is paramount. The primary solution, other than increased engine efficiency and ...

work page doi:10.3390/polym14225007 2021
[2]

Fabrication and testing of polymer-based capacitive micromachined ultrasound transducers for medical imaging,

C. D. Gerardo, E. Cretu, and R. Rohling, “Fabrication and testing of polymer-based capacitive micromachined ultrasound transducers for medical imaging,” Microsyst. Nanoeng., vol. 4, no. 1, p. 19, Dec. 2018, doi: 10.1038/s41378-018-0022-5. [25] J. Welsch, E. Cretu, R. Rohling, and C. D. Gerardo, “Ultrathin, High Sensitivity Polymer-based Capacitive Microma...

work page doi:10.1038/s41378-018-0022-5 2018

[1] [1]

Global Market Forecast 2025,

Introduction The aerospace industry is projected to experience unprecedented growth, with Airbus and Boeing each forecasting the delivery of over 43,000 new aircraft within the next two decades. [1,2] To limit environmental impact and operating costs, reducing fuel consumption is paramount. The primary solution, other than increased engine efficiency and ...

work page doi:10.3390/polym14225007 2021

[2] [2]

Fabrication and testing of polymer-based capacitive micromachined ultrasound transducers for medical imaging,

C. D. Gerardo, E. Cretu, and R. Rohling, “Fabrication and testing of polymer-based capacitive micromachined ultrasound transducers for medical imaging,” Microsyst. Nanoeng., vol. 4, no. 1, p. 19, Dec. 2018, doi: 10.1038/s41378-018-0022-5. [25] J. Welsch, E. Cretu, R. Rohling, and C. D. Gerardo, “Ultrathin, High Sensitivity Polymer-based Capacitive Microma...

work page doi:10.1038/s41378-018-0022-5 2018