TY - JOUR
T1 - The hydraulic characteristics of Roman lead water pipes
T2 - an experimental investigation
AU - Crapper, Martin
AU - Motta, Davide
AU - Sinclair, Coree
AU - Cole, Dominic
AU - Monteleone, Maria
AU - Cosheril, Adam
AU - Tree, Jonathan
AU - Parkin, Andrew
PY - 2022
Y1 - 2022
N2 - An experimental investigation was conducted on two lengths of Roman lead water pipe excavated near Corbridge, Northumberland, England. The pipes date from approximately AD 80. One length of pipe contained a sleeve joint. The pipes were linked to a pump system using 3 D-printed connectors, and the hydraulic head loss along the pipe was measured. The pipe without a joint was used to determine a value for the wall roughness height, fitting the results to the Darcy-Weisbach equation. The wall roughness height obtained was assumed to apply to the pipe with the joint, and the additional head loss observed was then assumed to be due to the joint, allowing determination of a local loss coefficient. The pipes have a sinter encrustation, indicating that they had typically flowed partially full. Using this as an indication of water depth and deriving topographical information from the excavation report, the likely flow the pipes carried during their operational life was estimated. It was concluded that the pipe wall roughness coefficient ks was 0.9 mm, the joint local loss coefficient was 1.159, and that during operation the pipe probably carried around 17 litres/minute.
AB - An experimental investigation was conducted on two lengths of Roman lead water pipe excavated near Corbridge, Northumberland, England. The pipes date from approximately AD 80. One length of pipe contained a sleeve joint. The pipes were linked to a pump system using 3 D-printed connectors, and the hydraulic head loss along the pipe was measured. The pipe without a joint was used to determine a value for the wall roughness height, fitting the results to the Darcy-Weisbach equation. The wall roughness height obtained was assumed to apply to the pipe with the joint, and the additional head loss observed was then assumed to be due to the joint, allowing determination of a local loss coefficient. The pipes have a sinter encrustation, indicating that they had typically flowed partially full. Using this as an indication of water depth and deriving topographical information from the excavation report, the likely flow the pipes carried during their operational life was estimated. It was concluded that the pipe wall roughness coefficient ks was 0.9 mm, the joint local loss coefficient was 1.159, and that during operation the pipe probably carried around 17 litres/minute.
KW - Roman
KW - pipe
KW - friction
KW - coefficient
KW - joint
KW - head losses
UR - http://www.scopus.com/inward/record.url?scp=85128779905&partnerID=8YFLogxK
U2 - 10.1080/17581206.2022.2054395
DO - 10.1080/17581206.2022.2054395
M3 - Article
SN - 1758-1206
VL - 91
SP - 119
EP - 134
JO - International Journal for the History of Engineering and Technology
JF - International Journal for the History of Engineering and Technology
IS - 2
ER -