﻿510 
  

  

  Collapse 
  of 
  Tubes 
  by 
  External 
  Pressure. 
  

  

  to 
  the 
  importance 
  of 
  the 
  yield-point 
  as 
  a 
  criterion 
  of 
  collapse 
  *, 
  

   In 
  the 
  light 
  of 
  the 
  above 
  discussion 
  we 
  can 
  see 
  why 
  the 
  

   group-averages 
  of 
  Professor 
  Carman's 
  results, 
  as 
  shown 
  in 
  

   fig. 
  3, 
  should 
  be 
  so 
  irregularly 
  distributed 
  ; 
  for 
  his 
  analyses 
  

   show 
  a 
  percentage 
  of 
  carbon 
  in 
  his 
  seamless 
  steel 
  tubes 
  

   ranging 
  from 
  017 
  to 
  0"22 
  ? 
  and 
  this 
  variation 
  cannot 
  fail 
  

   to 
  affect 
  the 
  yield-point. 
  

  

  Being 
  at 
  present 
  engaged 
  upon 
  an 
  experimental 
  investi- 
  

   gation 
  into 
  the 
  influence 
  of 
  length 
  upon 
  the 
  strength 
  of 
  

   boiler-flues, 
  I 
  have 
  thought 
  it 
  worth 
  while 
  to 
  employ 
  my 
  

   apparatus 
  for 
  testing 
  the 
  conclusions 
  of 
  this 
  paper. 
  Steel 
  

   tubes 
  of 
  the 
  same 
  thickness 
  and 
  diameter, 
  both 
  annealed 
  and 
  

   as 
  received 
  from 
  the 
  mills, 
  and 
  in 
  several 
  lengths, 
  were 
  

   tested, 
  and 
  the 
  results 
  are 
  given 
  in 
  the 
  following 
  table* 
  

   Care 
  was 
  taken 
  to 
  see, 
  by 
  measurement 
  with 
  a 
  micrometer, 
  

   that 
  the 
  process 
  of 
  annealing 
  had 
  not 
  affected 
  the 
  circularity 
  

   of 
  the 
  tubes, 
  and 
  the 
  comparative 
  weakness 
  of 
  the 
  annealed 
  

   specimens 
  may 
  therefore 
  be 
  attributed 
  solely 
  to 
  the 
  lowering 
  

   of 
  the 
  elastic 
  limit. 
  The 
  tubes 
  (1 
  inch 
  external 
  diameter 
  ; 
  

   0*028 
  inch 
  thick) 
  were 
  tested 
  under 
  hydraulic 
  pressure, 
  the 
  

   end 
  thrust 
  due 
  to 
  this 
  pressure 
  being 
  left 
  unbalanced, 
  and 
  

   the 
  results 
  are 
  shown 
  in 
  the 
  appended 
  table. 
  

  

  Comparative 
  Strengths 
  of 
  Annealed 
  and 
  Unannealed 
  

   Tubing. 
  

  

  Length 
  

   of 
  

  

  No. 
  of 
  Tests. 
  

  

  

  

  ■ 
  

  

  

  

  Tube 
  

  

  between 
  

  

  End 
  

  

  n6 
  

   to 
  

  

  T3 
  

  

  QJ 
  

  

  Is 
  

  

  Collapsing 
  Pressure 
  

   of 
  Annealed 
  Tube. 
  

  

  Collapsing 
  Pressure 
  

   of 
  Unannealed 
  Tube. 
  

  

  Strength 
  

   Ratio. 
  

  

  Plugs. 
  

  

  (Inches.) 
  

  

  

  a 
  

  

  

  

  

  869 
  

  

  1 
  

  

  1 
  

  

  1008 
  lbs. 
  per 
  sq. 
  in. 
  

   (mean). 
  

  

  1270 
  lbs. 
  per 
  sq. 
  in. 
  

   (mean). 
  

  

  1-26 
  

  

  6-19 
  

  

  3 
  

  

  2 
  

  

  1255 
  „ 
  „ 
  „ 
  

  

  1575 
  „ 
  ,, 
  

  

  125 
  

  

  4-19 
  

  

  2 
  

  

  1 
  

  

  1827 
  „ 
  „ 
  „ 
  

  

  2400 
  „ 
  „ 
  „ 
  

  

  1-31 
  

  

  It 
  will 
  be 
  seen 
  that 
  these 
  results 
  lend 
  strong 
  support 
  to 
  

   the 
  foregoing 
  theory, 
  which 
  discards 
  the 
  idea 
  of 
  any 
  differ- 
  

   ence 
  in 
  the 
  methods 
  of 
  failure 
  of 
  thick 
  and 
  of 
  thin 
  tubes, 
  at 
  

  

  * 
  Owing- 
  to 
  the 
  fact 
  that 
  the 
  compressive 
  " 
  hoop 
  " 
  stress, 
  in 
  fairly 
  

   thick 
  tubes, 
  is 
  not 
  distributed 
  quite 
  uniformly 
  across 
  a 
  section 
  of 
  the 
  

   tube 
  wall, 
  the 
  curve 
  of 
  fibre-stress 
  will 
  scarcely 
  be 
  so 
  flat 
  as 
  the 
  methods 
  

   of 
  this 
  naper 
  would 
  indicate. 
  

  

  