4 Saddle Vessel Design

Question # 1: I would like to have the t he guide lines for designing a vessel on multiple saddles? Answer: Guideline #1: "Avoid having more than 2 saddles!" Multiple saddles will be subject to diferential settlement ater which the load acting on the saddle and stressing the shell ma$ be greatl$ diferent rom the original design load% &essels beams on two supports ma$ e'perience diferential settlement but the loads will remain the same% Multiple saddles can be avoided b$ judicious location o the the saddles and use o wear plates and stifener rings% And wide saddles% Guideline #2: ( it is unavoidable to have three three or more saddles then the load distribution rom the shell to the saddle becomes a staticall$ indeterminate structural anal$sis problem% )ou will have to determine the load acting on each saddle using some method o structural anal$sis or continuous beams% *uggest then increasing these values b$ some "saet$ actor" to account or une'pected une'pected settlement etc some authors sa$ 2+,-.+,% /he anal$sis o the stresses in the shell at each saddle is then simpl$ perormed as or a vessel on a single set o saddles most li0el$ using the "ic0" or related *..++ method% Question # Question  # 2: (n order to ma0e use o e'isting oundations ( have to spread the load o a 1. m long 3%4 m diameter hori5ontal pressure vessel through three saddles instead o conventional two saddles% 6ould an$ one suggest a boo0 or a program programme me to deal with this7 Answer # 1: 6alculate the support reactions using beam theor$% theor$% /hen do a conventional conventio nal ic0 anal$sis but setting the load on the saddle e8ual to the calculated reaction% )ou will need to slot holes in two o the three saddles% Answer # 2: An$ urther oundation settlement will upset the load distribution predicted b$ structural anal$sis anal$sis as a continuous beam% A wa$ to minimi5e this efect would be to use onl$ two o the saddles%%% assuming assuming $our oundation is suitable or this%  /he 9t o the vessel vessel onto the saddles will will greatl$ greatl$ afect the the stress stress distribution distribution too% /he previousl$ previousl$ mentioned stress stress anal$sis oten cannot cannot cope with an ill-9tting saddle% )ou ma$ wish to consider welding the saddles to the vessel rather than the loose 9tting t$pe%  /he short is that that $ou cannot predict predict the the load at the three support support points with with great great con9dence due to non-A*M non-A*M 6odes issues% ( thereore thereore suggest that $ou add a health$ margin to the predicted predicted design loads% ( have oten used a 1;3 increase in load when aced with $our situation%/he A*M 6ode recommends onl$ two supports

under an A*M vessel or the reasons noted above%%% load prediction or design o the saddle supports%

Answer # 3:  /he oundations will most li0el$ e'perience some diferential settlement% A vessel on two saddles will not be greatl$ afected b$ this - the load on each saddle will remain essentiall$ the same% A vessel with three saddles is entirel$ diferent: ( the oundation on one end settles a bit more than the others A<< o the load now goes to the center saddle% /his triples $our load% ( $our e'isting oundation is a single solid slab $ou=ll be less susceptible to this% ( $ou have three individual oundations%%% Answer # > ?e had a customer that wanted to use . saddles on e'isting oundations to support an enormous @? hori5ontal pressure vessel that weighed much more than the pree'isting one because the$ had "done it that wa$ beore%" /he$ 8uic0l$ changed their minds when a structural engineer showed them how little sin0age o the oundation on one o the 9ve saddles could cause catostrophic damage with the much larger vessel% (t also helped that their old oundation wouldn=t handle the operating weight o the & period% @eedless to sa$ we shipped the & with 2 saddles and the$ built a new oundation% : *ome helpul resources: ressure &essel Bandboo0 - %C% Meg$es$ 12th edition 6odecalc sotware or pressure vessel desgin% (t will do saddle calculations based on supplied values% &er$ useul% www%coade%com Answer # .  /here is a stor$ which ( thin0 comes rom Galileo% Denaissance marble columns waiting to be installed were stored hori5ontall$ supported at both ends and sometimes bro0e% As these columns were a costl$ investment a consensus opinion was made that the$ should be supported both at the ends and also in the middle%  /his was done and shortl$ thereater all those supported at three points bro0e% ven .++ $ears later we are still able to all in the same trap% Question # 3: (=ve done a ic0 and ound a saddle si5e that will support the weight o the vessel 1+4" (E 1>. /-/ 1" th0 6* onl$ using the 2 saddles% (=ve been reading and researching the orums or ideas;wa$s to use the ic0 method or getting a sound engineering design on a multiple saddle pressure vessel% /he desire or a multiple saddle vessel is two-old: 1 Fn one end o the vessel there is an autoclave door which adds a lot o weight and 2 distribution o the total weight or the oundation especiall$ at h$dro it will weigh over 1 million lbs% (=ve thought about ta0ing the worst case scenario where loading is concerned which would be on the end where the autoclave door e'ists and doing a ic0 between the saddle or the door and what would be the ne't closest saddle and adding the loads present or the total vehicle concerning that section% An$ suggestions7

rian ressure &essels and Autoclave *$stems www%mcabeeconstruction%com