Predicting Corrugated Box Compression Strength Using an Artificial Neural Network

Siripong Malasri, Prasanth Rayapati, Divya Kondeti, (doi: 10.23953/cloud.ijapt.21)


McKee formula has been widely used to predict the compression strength of corrugated boxes. An experimental verification, published in early 2015, showed the inaccuracy of the formula. McKee formula left out several important factors, including box height, temperature, and humidity. An artificial neural network, CBU-BOX1, was developed based on 74 cases of cubical RSC single-wall corrugated boxes from 3”x3”x3” to 36”x36”x36”. Box height, temperature and humidity data were included in the network development. CBU-BOX1 performance ranged from 0% to 26.3% error with an average error of 6.9% while McKee formula performance ranged from 0.6% to 149.3% error with an average error of 28.7%. However, CBU-BOX1 performance dropped significantly when it was used for rectangular boxes. Out from twelve test cases of rectangular boxes, the formula resulted in an average error of 25.7% while CBU-BOX1 resulted in 34.8%. Thus, the network is unacceptable for rectangular boxes. In this study, 67 more cases were added to the previous 74 cases. Out of 141 cases, 43 were rectangular boxes. CBU-BOX2 significantly outperformed McKee formula with an average error of 9.21% versus 30.79%.


Artificial Neural Network; Compression Strength; McKee Formula; Corrugated Boxes

Full Text: PDF


  • There are currently no refbacks.

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.