Master thesis – Acoustic Leak Detection Signal Analysis

Every day, we get opportunities to make a positive impact – on our colleagues, partners, customers and society. Together, we’re pioneering the solutions of the future and unlocking the full potential of precious resources. Trusted to act on initiative, we challenge conventional thinking to develop world-leading technologies that inspire progress in vital areas, including energy, food, water and shipping.
As we push forward, the innovative, open spirit that fuels our 140-year-old start-up culture and rapid growth also drives our personal growth. So, as we shape a more resourceful, less wasteful world, we build our careers too.

We are looking for an M.Sc. thesis student to our Materials Technology and Chemistry department, R&D. So, what are you going to do today?

Alfa Laval

Alfa Laval is a global company with three key product areas: heat transfer, separation, and fluid handling. We have sales in more than 100 countries and in 2022 the order intake was about 5 500 million euro. The number of employees is about 20 000. The head office is in Lund, Sweden, where a major part of the research and development of plate heat exchangers reside.

The Master thesis project

Alfa Laval has developed a wide range of advanced heat transfer solutions for the transition to a cleaner energy future. One of the key quality factors of the heat exchanger is its leak tightness. A leak tight unit ensures the two medias from intermixing and leaking out to the surrounding environment. Leakages can occur both during manufacturing and in operation. The leakage root causes are numerous; Corrosion, liquid metal embrittlement, mechanical cracking etc.

To generate the next generation leak detection method Alfa Laval has purchased high fidelity recording equipment to capture the sound from the leakages. The aim of this project is to detect leaks from sound data recordings using features from time-frequency images, e.g. the spectrogram. 

The project includes:

• Recording of sound data with detected leakages together with an experienced laboratory technician and materials specialist.

• Analysis of the sound data recordings using various time-frequency visualization techniques.

• Suggestion of machine learning models suitable for extracting information from time-frequency images.

• Performance evaluation. 

Who are you?

We look for one or two curious and methodological student(s) with an interest in industry applications who wants to contribute to a sustainable future. Requirements: A course in stationary stochastic processes and knowledge in spectral analysis.

Assessment practicalities

We care about diversity, inclusion, and equity in our recruitment processes. We also believe behavioral traits can provide important insights into a candidate's fit to a role. To help us achieve this, we apply Pymetrics assessments, and upon application you will be invited to play the assessment games.

For more information about the project contact:

Axel Knutsson, Materials Specialist

[email protected]

Maria Sandsten, Mathematical Statistics, LTH,

marí[email protected]

Princess Jeremy Bondoc, Talent Acquisition Coordinator

[email protected]

This master thesis is planned to start no later than January 2025. We don't accept applications sent via email.