Ductwork pressure loss determination utilizing building information model

Con en om his wo k may be used unde he e ms o heC ea i eCommonsA ibu ion 3.0 licence. Any u he dis ibu ion o his wo k mus main ain a ibu ion o he au ho (s) and he i le o he wo k, jou nal ci a ion and DOI. Published unde licence by IOP Publishing L d Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 1 Duc wo k P essu e Loss De e mina ion U ilizing Building In o ma ion Model Michal No áček 1, Jose Remeš 1 1 Ins i u e o Compu e Aided Enginee ing and Compu e Science, Facul y o Ci il Enginee ing, B no Uni e si y o Technology, Ve eří 95, 602 00 B no, Czech Republic xpno ac03@ u b .cz Abs ac . Building design c ea ed as an in o ma ion model con ains no only in o ma ion o ma ed du ing building geome y c ea ion bu also addi ional in o ma ion illed in by a designe . The da a can be used u he as well as applied in building geome y analyses. One o hese a e analyses o p essu e loss in ai -handling uni s. Such analyses ake he ad an age o bo h c ea ed geome ic pa ame e s, and basic physical a iable. These analyses can be ca ied ou di ec ly in BIM so wa e wi hou ex e nal calcula ions needed. The main aim o his a icle is o desc ibe u iliza ion o in o ma ion in building model leading o ob ain p essu e loss in ai - handling uni s. 1. In oduc ion Building c ea ed as a building in o ma ion model p o ides plen y o oppo uni ies. In o ma ion con ained in model a e as comp ehensi e as he designe is. The model con ains undamen al geome ic in o ma ion which become he s anda d in ci il enginee ing. On he o he hand, building in o ma ion model can con ain complemen a y in o ma ion. These complemen a y in o ma ion a e especially physic o igin and can be u ilized o building analysing which is impo an o sus ainable building design. P ojec design phase is a building li e cycle phase when he mos impo an decisions a e made, and changes occu . The decisions a e impo an o u he ealiza ion and building using phase especially because o expendi u e. O e ime decisions con e ge o an op imal solu ion. In some cases, he building can be analysed di ec ly in o he modelling so wa e wi h no ex e nal applica ions o o he me hods needed. Comme cial so wa e al eady p o ides analysing ools u ilizing physical models. Howe e , mos o he analyses a e no possible o be u ilized di ec ly in modelling so wa e and one o hem is en ila ion duc wo k p essu e loss analysis. The analysing ools a e in luenced by building na ional s anda ds equi emen s. Al hough physical equa ions a e gene al, i s in e p e a ion and u ilizing depends on na ional habi s and s anda ds. Tha is he eason why inconsis ency is in he analysing ools be ween coun ies. The solu ion can be o de elop na ional s anda d analysing ools wi h he use o BIM da a in building design phase. [1] 2. Ven ila ion duc wo k p essu e loss One o he equi emen s o p ope en ila ion duc wo k design is a an powe su icien o supply ai in o e e y oom. [2] Duc wo k p essu e loss is needed o be calcula ed o designing he an. P essu e Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 2 losses a e caused by ai ic ion agains he duc walls [3]. The p essu e loss on s aigh duc sec ions is e y small. On he o he hand, in places whe e low di ec ion is changed (elbows, educ ions, o he en ila ion i ings), he p essu e loss is conside able. Fo p essu e losses in duc i ings (excluding s aigh pieces o duc ) we use he e m local p essu e loss. Local p essu e losses a e exp essed as a p essu e d op by he di e ence o s a ic p essu es a he beginning o he i ing and a i s end. This is based on Be noulli’s equa ion (1) whe e ρ is he ai densi y, is he mean eloci y in a gi en c oss-sec ion, p is he s a ic p essu e, h is he heigh di e ence be ween he measu ed c oss-sec ion and he e e ence plane and pz is he p essu e loss caused by ic ion. I is assumed ha he sys em is ai igh , and he mass low is he e o e cons an . I he duc c oss- sec ional a ea is cons an , he mean low eloci ies in di e en pa s o he duc a e equal. In he case o ai duc s, he e ec o g a i y on he low is o en neglec ed. Wi h his simpli ica ion, he p essu e d op o he duc i ing can hen be exp essed as equa ion (2) This p essu e d op can also be exp essed in e ms o mean eloci y as (3) whe e ρ is he ai densi y, is he low mean eloci y and ξ is he p essu e loss coe icien . The p essu e loss coe icien is based on he duc i ing ype and depends especially on he na u e o he low (Reynolds numbe ) and he i ing geome y. The p essu e loss coe icien alues can be ound in ables [3-5] o (less o en) in he o m o ma hema ical equa ions [3, 6-7]. Fo p essu e loss calcula ion on s aigh duc s sec ions is used equa ion (4) whe e λ is he ic ion ac o , l is he leng h o he s aigh duc sec ion, D is he duc hyd aulic diame e , ρ is he ai densi y, is he low mean eloci y. The Coleb ook-Whi e equa ion is o en used o he ic ion ac o calcula ion (5) whe e ε is he absolu e oughness o he duc ma e ial, D is he duc inne diame e and Re is he Reynolds numbe . In addi ion o he Coleb ook-Whi e equa ion, i s app oxima ion equa ions a e also o en used. These equa ions exp ess he ic ion ac o explici ly [6]. To selec he su icien an powe , i is necessa y o calcula e he o al duc wo k p essu e loss. Tha means o sum he p essu e losses o all i ing ypes and s aigh duc sec ions on he duc wo k. The designe is o ced o conside a ious geome ic pa ame e s and ypes o indi idual i ings o selec 𝜌𝑣1 2 2+ 𝑝1+ ρgℎ = 𝜌 𝑣2 2 2+ 𝑝2+ ρgℎ + 𝑝𝑧, 𝑝𝑧= 𝑝2− 𝑝1 ∆𝑝 = 𝜉𝜌 𝑣2 2, ∆𝑝 = $𝜆 ∙ 𝑙 𝐷∙𝜌𝑣 2 2, 1 √𝜆=%−2log+2,51 𝑅𝑒√𝜆+𝜀 3,7%𝐷5, Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 3 he app op ia e p essu e loss coe icien . Howe e , his p ocess can be au oma ed u ilized a duc wo k model in building in o ma ion model. 3. Building in o ma ion model Plen y o so wa e ha p o ides modelling buildings as BIM, also allows o model building se ices. The me e geome y o a gi en duc sec ion exp essed by pa ame ic alues can se e as an inpu a iable in he calcula ion o he p essu e loss. Some so wa e, such as Au odesk Re i , allow o analyse en ila ion duc s di ec ly in he applica ion’s use in e ace. Using men ioned so wa e, i is possible o analyse he en ila ion duc s o p essu e loss by means o se e al me hods. Fo s aigh duc sec ions i is possible o choose om a couple o app oxima ion equa ions (speci ically he Haaland equa ion and he Al shul-Tsaal equa ion) and he Coleb ook-Whi e equa ion (5) o ob ain he ic ion ac o . The in e nal calcula ion also conside s he na u e o he low acco ding o he alue o Reynolds numbe (so wa e dis inguishes be ween lamina and u bulen low). The o al s aigh duc sec ions p essu e loss is de e mined using he modelled geome y and speci ied amoun o supplied ai . The duc i ing p essu e loss is calcula ed acco ding o equa ion (3) u ilizing he mass low and he p essu e loss coe icien . The p essu e loss coe icien is ob ained in se e al possible ways: by he inpu o use -en e ed coe icien alue o each en i y (each duc i ing) o by he use duc i ing p essu e loss inpu di ec ly o each en i y and o by ob aining he coe icien alue om ASHRAE able [8] which is au oma ically assigned o each en i y acco ding o he i ing ype. Au oma ic i ing de ec ion in Re i and inding he p essu e loss coe icien in he ASHRAE able may seem like he bes choice o he designe . Bu his me hod has some disad an ages. The i s is he inabili y o edi he able (inabili y o edi he coe icien alues and pa ame e s) [9] and he second is he impe ec sea ching o duc i ing ype. The able consis s o geome ic pa ame e s o a de ined duc i ing ype, acco ding o which Re i inds he app op ia e coe icien alue. The p oblem occu s when he geome ic pa ame e s o he duc i ing do no co espond exac ly o he abula alues o he geome ic pa ame e s. In Czechia, duc i ings a e o en made- o-measu e and ha is he eason why he e is no app op ia e alue in ASHRAE able. In hese cases, he able is unable o in e pola e in e media e pa ame e alues. The exac ype o i ing is o en di icul o de ec ( o example, he di e ence be ween smoo h adius elbow and mi e ed elbow). The solu ion o hese impe ec ions can be ee access o abula alues and he possibili y o edi ing hem. The abili y o in e pola ion is also needed o he u ilizing o ma hema ical equa ions exp essing coe icien alues depending on geome ic pa ame e s and mo e accu a e de ec ion o duc i ing ype. 4. Complex p essu e loss solu ion in BIM The da a use om he building in o ma ion model o he pu pose o de e mining he duc wo k p essu e loss can be u ilized by se e al ways. When using comme cial so wa e, in o ma ion can o en be accessed wi h an ex e nal applica ion using API (Applica ion P og amming In e ace). As a esul , plen y o pa ame e s ha a e pa o he so wa e in o ma ion model can be u ilized as inpu a iables o he physical models c ea ed in he sou ce code o he ex e nal applica ion. An al e na i e o c ea ing an ex e nal applica ion may be o p ocess IFC ile. IFC ile is an in e changeable o ma con aining he undamen al pa ame e s o he in o ma ion model. This ile con ains numbe o pa ame e s ha can be used o analyse duc i ing p essu e loss. The ad an age is he e sa ili y o his ile. Comme cial BIM ools can wo k wi h his ype o ile. Howe e , he Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 4 disad an age is he limi a ion in he numbe o pa ame e s con ained in he ile. In pa icula , he physical pa ame e s wi h he duc i ing ype a e missing. In bo h cases, i is possible o use geome ic pa ame e s and p ocess hem u he . The main pa o he analy ical ool is i s physical model (physical equa ions), acco ding o which he calcula ion uns. In he case o p essu e loss, he key ac o is local p essu e loss, i s alues o en a y in he li e a u e. They a e mos o en w i en in abula o m. These coe icien alues a e cu en ly o en ob ained by nume ical me hods, bu wi h some simpli ica ions hey may di e om he ac ual measu ed alues on a eal duc wo k. Tabula alues may be unsui able o use in au oma ed calcula ion, as hei alues need o be in e pola ed and he e is a isk ha he en e ed pa ame e s will be ou o he ange o alues in he able. In such a case, i seems mo e app op ia e o use app oxima ion equa ions desc ibing he dependences o he magni ude o he coe icien o local p essu e losses on he geome ic pa ame e s o he duc i ing. Fo many duc i ings, he app oxima ion equa ions ha e no ye been de e mined o a e de e mined based on nume ical analyses, no eal duc wo k expe imen s. 5. Resul s and discussions Labo a o y expe imen o de e mine p essu e losses on he duc wo k is ime consuming bu co esponds mos o he eal s a e o he duc wo k in he building. F om he alues measu ed in his way, i is possible o compile equa ions, which can be used in he au oma ion o p essu e loss calcula ions in he building in o ma ion model. An example o an au oma ic p essu e loss calcula ion could be an ex e nal applica ion / add-on o Re i ha calcula es p essu e losses o undamen al duc i ings, which a e elbows and ansi ions o ound and ec angula pipe c oss-sec ions. On he in e ne , his applica ion called Re i Duc wo k P essu e Loss Calcula o is loca ed on he Gi Hub web hos ing in he o m o an open-sou ce add-on o he Re i applica ion. I is an ex e nally accessing applica ion ha uses he in e nal pa ame e s o he Re i so wa e o de e mine he ype o duc i ing and hen assigns he app op ia e p essu e loss coe icien o each i ing u ilizing geome ic pa ame e s. The coe icien is assigned om wo di e en sou ces. One sou ce is a abula alue acco ding o he ASHRAE able om 2001 and he o he is ma hema ical equa ion e lec ing he dependence o he coe icien on he geome y. Fo hese pu poses, he labo a o y expe imen esea ch es ablished analy ical equa ions depending he coe icien o local p essu e losses on 45° and 90° elbows geome y pa ame e s. The equa ions a e inco po a ed in o he applica ion and can be ound in he Table 1. The ad an age o he applica ion is he possibili y o i s use edi ing. Edi ing abula alues, including he alues o hei pa ame e s, is use - iendly wi hou he need o in e ene in he sou ce code, as hese alues a e s o ed in a ex ile. I is he e o e possible o adap he coe icien alues o designe expe ience o p e e ences. The men ioned applica ion has he po en ial in Re i so wa e o be de eloped in o an au oma ic ool o he analysis o p essu e losses o he Czech (and o in e na ional) en i onmen . The applica ion de elopmen consis s mainly in adding o he duc i ings o he sou ce code, which he algo i hm would conside and hus con ain a sa is ac o y numbe o i ings o e ime. O he elemen s sui able o inclusion in he calcula ion could also be he e ec o elbows ha a e placed downs eam nea ly each o he wi h small o no duc s aigh pa . This ad anced analysis can lead o mo e accu a e esul s o p essu e losses. Fo wo consecu i e elbows, he abula alues a e o en unsa is ac o y, and he analy ical exp ession o he coe icien -geome y dependency is no common. Fo his pu pose, ec angula c oss-sec ion elbows wi h di e en aspec a ios o c oss-sec ions and di e en alues o axial ounding and di e en sizes o he s aigh in e media e pa be ween he elbows we e analysed. The esul is he equa ion ha can be inco po a ed in o an analy ical ool ha u ilizes in o ma ion in a building model. The equa ions o p essu e loss dependence a e gi en in he ollowing Table 1. Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 5 Table 1. Dependency o p essu e loss coe icien on he geome y (W is wid h o he ec angula duc c oss-sec ion, H is heigh o he ec angula duc c oss-sec ion, L is leng h o in e media e s aigh sec ion). Duc wo k ype Dependency o p essu e loss coe icien and duc i ing geome y Elbow 1 × 45° Elbow 1 × 90° Elbows 2 × 45° Elbows 2 × 90° Table 2. Measu emen e o s, equa ion alidi y in e als, esidual sum o squa es and le el o signi icance (W is wid h o he ec angula duc c oss-sec ion, H is heigh o he ec angula duc c oss- sec ion, L is leng h o in e media e s aigh sec ion). Duc wo k ype Absolu e e o Validi y in e al Residual sum o squa es (RSS) Le el o signi icance α Elbow 1 × 45° 0,339 0,95 Elbow 1 × 90° 0,395 0,95 Elbows 2 × 45° 8,578 0,95 Elbows 2 × 90° 30,390 0,95 6. Conclusions The building in o ma ion model s o es a subs an ial amoun o in o ma ion ha can po en ially be used. The mo e in o ma ion he model con ains, he mo e p ac ical u ilizing ways he e a e. U ilizing he BIM model o analy ical asks is an ad anced ea u e o design so wa e and o e s conside able po en ial, which allows us o make compe en decisions based on analysis and da a a a ime when we can in luence he shape o he building wi h a minimum o cos s. By simply using he al eady exis ing pa ame e s and supplemen ing he physical calcula ing, i is possible o au oma e and o s eamline building analysing. In his way, we can also maximize he use o a ailable in o ma ion in he model. In addi ion o he bene i apid ans e o all c ea ed in o ma ion abou he building, BIM also o e s he po en ial o applica ion in he physical cons uc ion ield. The duc wo k p essu e losses de e mina ion is necessa y o he design o en ila ion sys em in a building. This p ocess can be leng hy and complica ed. The undamen al a iables o he calcula ion, which a e he coe icien s, o en di e and a e no uni o m o he Czech en i onmen . In his espec , 𝜉 =−12,970*𝑊 𝑅-2+31,350𝑊 𝑅−18,490 𝜉 =−5,606(𝑊 𝑅+2+14,210𝑊 𝑅−8,470 𝜉 =0,043'𝐻 𝑊*−2,441 −0,015'𝐿 𝑊*3+0,072'𝐿 𝑊*2 𝜉 =0,432(𝐻 𝑊+−1,248 −0,007(𝐿 𝑊+3+0,032(𝐿 𝑊+2 𝑊 𝑅= (1,0244; 1,3548) 𝑊 𝑅= (1,0244; 1,3548) 𝐻 𝑊= (0,254; 1,000) 𝐿 𝑊= (1,476; 3,952) 𝐻 𝑊=(0,254; 1,000) 𝐿 𝑊= (1,476; 3,952) Building De ec s 2020 IOP Con . Se ies: Ma e ials Science and Enginee ing 1066 (2021) 012011 IOP Publishing doi:10.1088/1757-899X/1066/1/012011 6 he uni ica ion o alues o ma hema ical dependencies can be a s ep owa ds s anda dizing his analysis and a simpli ying, e ining, and speeding i up. Wi h simple au oma ion p ocesses, we can sa e a conside able amoun o ime and hus design buildings e icien ly in he con ex o cu en needs. Re e ences [1] P. Smi h, “BIM implemen a ion - global s a egies,” P ocedia enginee ing, pp. 482-492, 2014. [2] P. Schild, and M. Mysen, “Recommenda ions on Speci ic Fan Powe (SFP) and Fan Sys em E iciency –IV,” TZB-in o, P ague: Jícha, 2014. [3] J. Chyský, and K. Hemzal, “Ven ila ion and ai condi ioning,” Thi d comple ely ewo ked edi ion. P ague: B p ess, B no; 1993. [4] I. Idelchik, D. G unae , and A. Ba ouch, “Handbook o Hyd aulic Resis ance,” X/5. Mosk a- Lening ad: Gosuda s ennoe Ene ge icheskoe Izda el's o, 1960. [5] K. Labou ka, and T. Suchánek, “Local p essu e loss coe icien alues – hea sou ces and undamen al duc i ings,” TZB-in o, P aha, ©2001-2019. [6] J. Schwa ze , and V. Zm hal, “De e mina ion o local p essu e loss coe icien s u ilizing,” A icle. 2010. [7] S. Genić, I. A andjelo ić, P. Kolendić, M. Ja ić, N. Budimi , and V. Genić, “A Re iew o Explici App oxima ions o Coleb ook’s Equa ion,” Re iew, 2011. [8] R. Howell, H. Saue , and W. Coad, “P inciples o Hea ing, Ven ila ing, and Ai Condi ioning,” Ilus a ed edi ion, A lan a: Ame ican Socie y o Hea ing, Re ige a ing and Ai -Condi ioning Enginee s, 2005. [9] Au odesk. I 's no possible o modi y ASHRAE Table in Re i . Au odesk Knowledge Ne wo k, Au odesk Inc.; 2017.