Video from my YouTube channel

Subscribe to My You tube channel

An Assessment of Artificial Intelligence Capabilities in Structural Analysis: A Comparative Study of Determinate and Indeterminate Beams (Research paper)


This paper critically evaluates the performance of several AI models in structural analysis. The capability of different AI systems, including ChatGPT, DeepSeek, Gemini, and Perplexity, was assessed using a series of beam problems with varying degrees of static indeterminacy and structural complexity. The outputs were benchmarked against classical analytical methods, namely the force method and slope-deflection method, as well as numerical results obtained from structural analysis software (SAP2000). The results indicate that AI models can produce reasonably accurate results for statically determinate beams and systems with low levels of indeterminacy. However, their accuracy decreases significantly as structural complexity increases, with notable discrepancies observed in highly indeterminate systems. The study concludes that current AI models lack the effectiveness required for reliable structural analysis of complex systems and should not be used by engineers as the obtained results are not reliable especially for complex and statically indeterminate structures.

Read full paper from here: 
Link
 


Comments

Videos from My Youtube channel

Popular posts from this blog

Rebar Splicing and Overlapping — ACI 318-19 Requirements Explained

stress-strain diagram for aluminium alloy and rubber

Binding wire for steel reinforcement

Pile cap

Non-Working Pile Testing and Static Compression Load Test

Prestressing stands debonding

Bridge Top slab

Fire resistive construction-firewalls

Nominal maximum size of coarse aggregate

Repairing of concrete bulging