CALCULATION THE ENVIRONMENTAL IMPACTS OF BLEVE USING ARTIFICIAL NEURAL NETWORKS

Barisik T., Guneri A. F.

FRESENIUS ENVIRONMENTAL BULLETIN, cilt.30, ss.9611-9616, 2021 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 30
  • Basım Tarihi: 2021
  • Dergi Adı: FRESENIUS ENVIRONMENTAL BULLETIN
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Aerospace Database, Aqualine, Aquatic Science & Fisheries Abstracts (ASFA), CAB Abstracts, Chemical Abstracts Core, Communication Abstracts, Environment Index, Geobase, Greenfile, Metadex, Pollution Abstracts, Veterinary Science Database, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.9611-9616
  • Anahtar Kelimeler: BLEVE, environmental impact, thermal radiation, fire, artificial neural networks, Levenberg-Marquardt algorithm, THERMAL-RADIATION, HAZARDS
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

For the community, major manufacturing facilities pose a high fire risk. BLEVE is known as the sudden outflow of a superheated liquid resulting from the liquid's evaporation under pressure in the atmosphere. The reason for these sudden outflows is caused by fires around the tank, corrosion and overheating inside the tank. Before using the modeling methods in the literature, artificial neural networks were modeled with input data determined using the Levenberg-Marquardt algorithm, which is a multi layer sensor (MLP) (teacher teaming) sort. The real values obtained from the measurement are compared to the network's outputs. The outputs of the real values. which produced, and of the outputs produced by the network using the real inputs are compared the compatibility. The feasibility of performance determination was investigated by using the neural network and curves formed from the resulting values. For BLEVE, the fireball explosion, a scenario was calculated. Artificial neural network results were generated for the outputs of the data on LPG gas explosion in tankers, which have 100, 150 and 200 m(3) volumes. Input values in the generated artificial neural network model; BLEVE's expansion measurement distance values. Thermal radiation heat flux (W/m(2)) values were predicted from the network as an output corresponding to these inputs. In addition, graphics comparing the actual expected values with the values found by the network created with the Levenberg-Marquardt Algorithm were also added.