Defect detection with a thermal camera

Abstract

The construction sector is undergoing substantial growth and there are many challenges to address. One challenge is building defects, which affect the structural integrity and put the structure at risk. A common and intrusive building defect is moisture, which can slowly expand and lead to complicated defect detection. Moisture intrusion can cause many issues in the structure of the building and poses a health risk to the occupants due to mold growth. It can be caused by various causes, some easily solvable, and others need professional help.

One of the most common moisture detections is by visual inspection. This inspection includes a visual walk through inspection and highly depends on the judgment of the inspector which is time-consuming and not reliable. Other methods which are used in the industry to detect moisture intrusion, include a pin-type moisture meter, gravimetric method, Karl Fischer Titration and many more. However, these methods are invasive and lead to other damages and lengthy defect detection processes. Nowadays, moisture intrusion can be detected by thermal cameras, they are non-intrusive and quick to detect any abnormalities, however, there is a lack of knowledge on the effectiveness of thermal cameras when detecting defects. This study aims to investigate the effectiveness of thermal cameras to detect moisture defects within building structures with a qualitative method. By exploring the objectives of the study, which include the benefits and limitations of thermal cameras, investigation of the accuracy of thermal cameras by conducting experiments, comparison of the data from qualitative results with software simulations, the influence of temperature and weather on thermal images and lastly recommendations for future work.

The methodology consists of a literature review, experimental analysis and simulation analysis. Experiments were conducted in the Energy Building Design (EBD) laboratory at Lund University, where moisture was introduced to already existing walls and monitored with thermal cameras and moisture meters. The experiments were conducted on both exterior and interior walls to analyse the influence of the environment on thermal images. The study includes simulation performed by WUFI software, with materials and moisture intrusion volume similar to the experimental properties, making the data comparable. By exploring the benefits, limitations and study development of thermal cameras there were boundaries of expectations drawn which helped to guide the experiments.

To investigate the accuracy of moisture defect detection, the experiments involved creating interior and exterior holes on the same wall. The purpose of the two holes of depth 6.8 cm and 10.6 cm on the interior wall, was to observe moisture transfer from interior to exterior surface. Two exterior holes with a depth of 1.5 cm and 2.5cm were made to measure at what depth the moisture occurs on the surface and how the moisture transfer is impacted by environmental conditions such as humidity, wind-driven rain and radiation.

As part of the experiments, water was inserted into the holes daily, increasing the volume from 25 ml to 100 ml on the interior wall and from 15 ml to 25 ml on the exterior wall. The Protimeter moisture content tool was used to measure the data on moisture expansion. Protimiter is a pin-type moisture meter which by creating electrical resistivity between pins measures the percentage of moisture equivalent. A FLIR E6 thermal camera was used to detect infrared radiation emitted by both sides of the wall. The thermal camera measured the inserted moisture from the indoor experiments on the outdoor surface. For the outdoor experiments, thermal images were made on the exterior. The thermal imaging and moisture content measurements were conducted daily. To keep track of the outdoor and indoor temperature, the HOBO U12-012 tool was used on both surfaces. This could be used to assess how much solar radiation impacted the outdoor surface.

The results show that the thermal camera effectively detected moisture defects. For the interior wall, moisture was detected for the hole with a depth of 10.6 cm on the outside surface. The moisture slowly expanded daily and was detectable by the thermal camera. Moisture in the 6.8 cm hole could not be detected, as it was too far from the exterior surface and the moisture did not penetrate through the materials. For the outdoor wall, moisture was detected for the hole with a depth of 1.5 cm on the outside surface. This was due to the materials that were close to the outermost layer that were not present in the 2.5 cm hole.

In addition to the experiments, hygrothermal simulations of the wall under several scenarios were carried out using WUFI software. The scenarios included a base case wall without added moisture and cases that represented the conditions in the experiments. The data from the simulations were compared with the daily experimental measurements.

To see if the software predicted the moisture intrusion similarly to experimental results, software simulation data were compared with qualitative results. The volume of moisture content in the simulation was significantly lower compared to the experimental data. This can be due to factors not included in the simulations, such as wind-driven rain and humidity. The water inserted in the experiment was more concentrated compared to the simulations and the location of the experimental moisture measurements was closer to the hole than the simulations.

Temperature and weather had a great influence on thermal images, especially on the outdoor wall which was constantly impacted by weather. The solar radiation on the outdoor surface caused reflections which led to false positive readings. Shadows on the outdoor surface caused the material to be colder which led to false readings. Additionally, false readings were caused by water droplets from rain which reflected the surrounding light.

Overall, this study shows that thermography is a partly effective tool to use when detecting moisture intrusion in buildings, however, it still has notable limitations which can influence the accuracy of the thermal readings, such as depth of moisture and environmental factors. The moisture defects which were located deeper in the structure were not detected and environmental factors such as solar radiation or rain did also create false positive and false negative results.

For any further studies and experiments, it would be needed to fully assess the thermal camera settings and use variable materials and depth of moisture intrusion. Additionally, it would be interesting to extend the experimental period and measure data at various points during a day with various weather conditions. Another recommendation would be to start the experiment with a very small volume of water, prolong the experimental period and slowly add more water, measuring the data more often.

Popular Abstract

The construction sector faces many challenges daily, including moisture defects, which can affect structural integrity. Moisture defects are often well-hidden, can cause mold growth which is especially difficult to detect and poses a health risk to the occupants. Moisture can infiltrate buildings through condensation, external leaks or rising dampness. Detecting moisture intrusion with traditional methods is a time-consuming process and typically intrusive, causing damage to the building. A traditional non-invasive method is visual inspection, which is not very reliable. Nowadays, moisture intrusion can be detected by thermal cameras, a non-invasive and quick procedure that detects abnormalities in temperature, indicating moisture. However, thermal cameras have limitations that can influence thermal imaging and are easily misinterpreted. These limitations are weather, influence of temperature and various material properties.

This study aims to investigate the effectiveness of thermal cameras in detecting moisture defects, by conducting a literature review and analysing the benefits and limitations of thermal cameras through conducting experiments and comparing with software simulations. Experiments were done in the Energy Building Design laboratory at Lund University, in a controlled environment where moisture was inserted into interior and exterior walls daily and results were monitored using a thermal camera and moisture meter (pin-type method). Experimental results were later compared to the simulated data. The findings showed that thermal cameras can detect moisture that is close to the surface, however, thermal cameras cannot detect moisture deeper in the structure. Simulation data using WUFI highlights the importance of using simulations alongside physical experiments, as simulations perform predictive long-term performance analysis which provides early design feedback and experiments can be adjusted accordingly.

Thermal cameras show potential for detecting moisture intrusion in buildings due to being a quick and non-invasive method, however, many limitations influence the accuracy of readings. During the experiments, the depth of the defect and reflection of sunlight were observed to contribute to false readings.

For future research, it would be recommended to prolong experiments over a longer period and adjust the volume of water inserted, optimize both camera and pin-type meter settings, and test the camera with various materials to see how it behaves and understand the readings. Additionally, inserting a smaller amount of water daily and monitoring the experiments more frequently (morning and evening) could lead to valuable insights. This would provide an extensive assessment of moisture detection using thermal cameras and how well the defect can be detected over time. It would be beneficial to conduct WUFI simulations before the experiments, as it might offer valuable insights into the length of the experimental period.

• author: Krupová, Karin
• supervisor: Pieter de Wilde; Niko Gentile
• organization: Division of Energy and Building Design
• course: AEBM01 20241
• year: 2024
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: Thermal camera, Defect detection, Moisture intrusion
• language: English
• id: 9176281
• date added to LUP: 2024-10-08 12:45:14
• date last changed: 2024-10-08 12:45:14