Octane sensitivity (OS), defined as the difference between research octane number (RON) and motor octane number (MON), is a topic of significant interest to the engine development community due to its apparent effect on knocking conditions. Fuels with a high octane sensitivity have been shown to enable higher efficiencies by allowing for elevated compression ratios and the implementation of turbocharging. In practice, RON and MON must be experimentally tested to determine OS; however, the experimental protocols require specialized equipment, a significant financial investment, and a substantial amount of time. To this end, predictive computational models, based on molecular properties, would allow researchers to preemptively screen compounds of interest prior to experimental testing. This work presents a model that uses artificial neural networks (ANNs) trained with quantitative structure-property relationship (QSPR) descriptors to predict RON, MON, and OS. Additionally, the viability of deriving OS from RON and MON predictions (defined as derived sensitivity, dOS) is evaluated against the direct OS predictions. Models for RON and MON are shown to have root-mean-square errors (RMSEs) of 7.717 and 8.119 respectively. Additionally, dOS predictions are shown to be more accurate than OS predictions with RMSEs of 6.784 and 7.697 respectively, showing it is more beneficial to obtain OS predictions from RON/MON predictions than predicting OS directly. Furthermore, individual QSPR descriptor relationships to RON/MON/OS are discussed, highlighting the correlation between key structural features and RON/MON/OS.

Predicting RON, MON and octane sensitivity using artificial neural networks

Macromolecular Chemistry: The Second Century

[Author Benefits](https://scimeetings.acs.org/?utm_source=pubs_content_marketing&utm_medium=website&utm_campaign=0320_MCF_NPI_Launch_Spring_Homepage&ref=pubs_content_marketing)

[How to use SciMeetings](https://storage.googleapis.com/dl.morressier.com/acs/SciMeetings-ACSSpring2020-HowToGuide_M3020-Final.pdf)

[Frequently Asked Questions](https://storage.googleapis.com/dl.morressier.com/acs/SciMtng_ACSNMSp2020_FAQ_M3120_Final.pdf)

ACS Spring 2020 National Meeting & Expo

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.