This master's thesis provides knowledge on non-destructive testing of cement and supplementary cementitious materials composition in concrete with a handheld X-ray Fluorescence analyser.

Today, concrete production is responsible for 8% of the world’s CO2 emissions. Recycling concrete material can directly contribute to the reduction of CO2 emissions. This process is costly and time-consuming. A possible solution would be to use a handheld X-Ray fluorescence spectrometer on-site to determine the concrete chemical composition.

No existing research indicates if concrete identification with a handheld X-Ray Fluorescence analyser is possible. This thesis intends to prove that this technique can differentiate concrete with various chemical composition.

Fourteen concrete cubes of fourteen different chemical compositions were analysed to fulfil this objective. The fourteen compositions reflect the concrete design used in the Netherlands. Experimental programs conducted on the concrete revealed the impact of different factors on the results obtained from the handheld-XRF. These factors include measurement time, moisture, surface carbonation, and matrix effect. Each factor impacts various oxides in different proportions, leading to distinct patterns. After investigating their impacts, a protocol was written to test all the mixes. Finally, the reproducibility of the protocol was assessed, and the mixes were tested using the protocol.

The primary outcome of this thesis is proof that twelve of the fourteen mixes were differentiated based on their alumina content. This oxide proved to be less impacted by moisture and surface carbonation than the other oxides. The influence of the different factors on measurements was identified and quantified. These studies also revealed that twenty measurements were sufficient to identify the mixes. The protocol improved the control of the factors but also appeared limited by the concrete matrix.

A possible approach to circumvent this problem would be considering oxides content as thresholds rather than numbers. Determining these thresholds requires testing many samples. Another further study is the possibility of reducing the impact of moisture and surface carbonation on-site.