The virtual container: Physics-based simulation of refrigerated container map temperature and fruit quality evolution and variability in a shipment

Thijs Defraeye*, Celine Verreydt, Julien Gonthier, Leo Lukasse, Paul Cronjé, Tarl Berry

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


Many fresh food and vegetables are transported in refrigerated containers after harvest, often over thousands of kilometers. A better understanding of when and where food quality is lost in these supply chains provides opportunities to reduce the quality variability within and between different shipments. Nowadays, however, only a few (hygro-)thermal sensors are placed within every shipment, which masks the variability in the shipment. These hygrothermal data are also not actionable for stakeholders for decision-making. The resulting food quality evolution and its variability within a shipment remain invisible. We approach this problem by building a validated physics-based digital model of a refrigerated container for citrus fruit. This virtual container model is described extensively in an accompanying paper (Defraeye et al., 2024). We use computational fluid dynamics with a two-phase porous media approach to simulate the airflow in this virtual container. We also simulate the cooling process of every single fruit and the fruit's thermal quality loss. We compare the virtual container model with a full-scale experimental data set. The simulations captured the main physical trends of container cooling but cooled on average 0.3 d faster. The variability in seven-eighths cooling time within the cargo was over 2 days, and that of the remaining shelf life after the transport period of 24 days was about 0.7 days. We identify the slowest cooling location in the cargo. This location is the pallet or box that would need to be inspected to assess the quality or the best location to place the sensors. The model simulations indicate that during the container's warm loading or hot stuffing, high airflow rates should be used for the first three days to improve fruit quality preservation. Lower airflow rates can be used later on. The simulations show that airflow bypasses through gaps between pallets should be avoided. Using a void plug can decrease the cooling time by 30%. Void plug placement is found to be much more effective than reducing the small gaps between the pallets. The type of void plug that is used is less critical. Cooling and quality problems could be mitigated by placing precooled pallets at the expected slowest cooling locations in the container. Changing the T-bar floor height, while keeping the pallet height constant, affected the differences in cooling and quality between both sides of the container. The virtual container provides a full spatiotemporal map of the fruit temperature, temperature-driven quality, and postharvest life for all fruit in the container. We quantified cooling times and remaining shelf life in 60,000 individual probe locations. The data that are engineered by the virtual container is currently a missing link to enable in-transit temperature management, shelf-life-driven logistics, and inventory management. The virtual container is also an essential building block of a refrigerated container's digital twin that can help reduce food loss and increase supply-chain resilience. Such simulation tools will support stakeholders in the future in evaluating and improving cargo temperature control and resulting fruit quality at arrival.

Original languageEnglish
Article number112722
JournalPostharvest Biology and Technology
Publication statusPublished - May 2024


  • Citrus
  • Computational
  • Digital twin
  • Multiphysics
  • Refrigeration: slot-ventilated enclosure


Dive into the research topics of 'The virtual container: Physics-based simulation of refrigerated container map temperature and fruit quality evolution and variability in a shipment'. Together they form a unique fingerprint.

Cite this