Nintendo Engineering: Where are We with Our Modeling?

Transcription

1 Nintendo Engineering: Where are We with Our Modeling? Dr. Achilles Karagiozis Distinguished Research & Development Staff Engineering Science and Technology Division May 25, 2006

2 Acknowledgments: BCBEC Moisture Group Incorporated

3 It all starts with the Theory : Perm Molecules 3

4 Presentation Roadmap Current Understanding New Standards Application Future Conclusions 4

5 Challenge Develop the scientific competencies to analyze complex dynamic heat, air and moisture transport processes in porous media (Hygrothermal) Develop a framework to evaluate & characterize the performance of building envelopes systems and building stock Use the scientific competencies & framework to develop guidelines for moisture control, improve Building Codes validate new innovative products & educate building designer & architects 5

6 Why have Envelopes worked well in the Past? 6

7 Past Approach Trial and Error Attention to Detail but Little Building Science Worked until: Enhanced Comfort Requirements Energy Conservation Material started to Change 7

8 What is needed is BETTER DESIGNS Increased drying performance Solution : DESIGN Better Water Management DESIGN More Forgiving Systems Innovative Materials 8

9 Building Science Approach PHYSICS NUMERICS Define Physics BUILDING SYSTEM SUB-SYSTEM Define Load Inputs Define Material Response BOUNDARY CONDITIONS MATERIAL PROPERTIES Define Construction Systems & Sub-Systems Systems 9

10 Models Development of a scientific model Problem Literature review Mathematical Model Software Optimization Experimental validation in the laboratory and field tests Scientific model 10

11 Models Old Approach Type Transport Phenomena - Heat conduction - Vapor diffusion - Heat conduction - Vapor diffusion - simplified capillary flow - Heat transport - Vapor diffusion - Heat transport - Vapor diffusion - Liquid transport - Heat transport - Convection - Heat transport - Vapor diffusion - Convection - Heat transport - Vapor diffusion - Convection - Heat transport - Vapor diffusion - Convection - Heat transport - Vapor diffusion - Liquid transport - Convection Steady state X X X X X X X X X X X Transient X X X X X X X X X X X X X X Number of Models IEA Annex 24 - Task 1 (Hugo Hens) 14 countries - 37 models 11

12 Models New Approach 12

13 Models 13

14 Out of the Date Models Glaser / Dew Point Method Compute the temperature and saturation vapor press. profile plot p versus diffusion resistance: risk of condensation Limits: only steady state behaviour only diffusion no heat and moisture storage no coupling of heat and moisture transfer 14

15 Models Authors Models History 1856 Darcy 1957 Krischer 1957 Philipp & de Vries 1958 Luikov 1983 Kießl 1987 Häupl & Stopp 1990 Rode MATCH 1992 Garrecht 1994 Künzel WUFI 1994 Karagiozis & Salonvaara LATENITE 1997 Grunewald DIM, DELPHIN 1999 Bednar 1999 Mendes UMIDOS 2001 Karagiozis MOISTURE-EXPERT 15

16 Trends of Software Developments building product manufacturers (Level I) practitioners (Level II) educational establishments (Level III) production production system prefabricated buildings design process quality control maintenance, operation universities carpenters, further bricklayers training of practitioners S U PP O RT F L E X I B I L I T Y 16

17 Models Construction Material properties H T Hygrothermal Model Moisture w φ = φ φ t T t = ( D φ + δ ( φ )) Energy p p sat ( ) ( λ T ) + h v δ ( φ p ) p sat Climate Boundary conditions Temperature Field Moisture Field Heat Fluxes Moisture Fluxes 17

18 Models Overview of available HAM-software Name of the model Authors Heat conductivit y Diffusion Capilarity MATCH Rode WUFI-ORNL/IBP Künzel, Karagiozis, WUFI-Pro 3.2 Künzel, Schmidt, Holm WUFI2d Künzel, Holm, Eitner Delphin Grunewald 1 und d-HAM Hagentoft, Blomberg ConDry Hedenblatt, Arfvidsson Umidos Mendes, Ridley Convection Dimensions Luftströmung Enthalphyptransport 18

19 Models The operator requires knowledge, skill, and experience Important to balance input data and results with engineering experience and judgement Must understand boundary conditions material properties transport mechanism deterioration/damage mechanism construction realities Most models are presently 1-D 19

20 Advanced Hygrothermal Modeling WUFI- ORNL WUFI-2D ORNL MOISTURE-EXPERT EXPERT ME-STOCH 20

21 Calculation of coupled Transport: Programming Material properties Initial conditions Climatic data Time steps Construction Numerical grid Input Surface transfer Control parameters New time step Update thermal coefficients Calculate temperature field Update hygric coefficients Calculate moisture field No convergence Yes Temperature fields Heat fluxes Output Moisture fields Moisture fluxes 21

22 The Science Transport Processes Complex during day time 22

23 The Science Transport Processes Complex during night time 23

24 Engineering vs Physics Analysis (REAL VERSUS IDEAL) MODEL MOISTURE-EXPERT (Karagiozis, 2001, 2004) Liquid Flow Governing Equations Moisture Balance The moisture transport balance is given as: ( ρ ( T ) u) m t ( u, T, x, y) φ δ ( u, T ) = D P + φ p v Liquid flow Vapour flow r ρ V v a { Air flow Vapor Flow Natural & Forced Convection 24

25 Engineering vs Physics Analysis (REAL VERSUS IDEAL) Mass Balance ρ a t ( T ) r +. = ( ρ ( T ) ν ) 0 a a Momentum Balance ( ρ ( T ) ν ) a t r a + ( ρ ( T ) ν ; ν ) a r a r a = P a + µ a K ( T ) a r ν a r + ρ ( T )g a Energy balance ρ m ( u, T ) C ( u, T ) =. ρ C ( T ) p T t r ( V T ) +. k( u, T ) a p a Convection ( P ) ( T ) + L. δ ( u, T ) ( T ) f l L. ρ ( u, T ) u ice m t Condensation Conduction v p v Evaporation 25

26 MOISTURE-EXPERT v.2.1.3a 2-D Capabilities Vapor Air Flow Vapor and Liquid Diffusion Solar and Sky Radiation Wind-Driven Rain Moisture-Thermal Sources and Sinks Dynamic Stack and HVAC Effects Temperature Dependent Sorption Processes 26

27 WUFI, WUFI-ORNL 1-D Capable Vapor Air Flow Vapor and Liquid Diffusion Solar and Sky Radiation Wind-Driven Rain Moisture-Thermal Sources and Sinks, 2006 Cladding Ventilation (2006) 27

28 Model Validation Very few models have been validated Different levels of validation exists Material level Laboratory level Field level The most validated hygrothermal model worldwide is WUFI, WUFI-ORNL The most validated research model worldwide is MOISTURE-EXPERT 28

29 Was the model Validated??? Yes.. Yes.. and Yes Real Field & Lab Data ASHRAE TRP 1091 PSU/UW/ORNL Seattle WSU/DOE/ORNL Project Charleston EIMA/DOE/ORNL 29

30 Laboratory Validation (0.8 Lps) Panel 8 - Weight Change % (E-S)/E <0.35 %> 2.6 % (E-S)/E <0.4 %> % (E-S)/E < 0.5 %> Weight change Moisture Expert 1000 Weight (g) % (E-S)/E <1%> % (E-S)/E < 4% > /3/02 12:00 AM 7/4/02 12:00 AM 7/5/02 12:00 AM 7/6/02 12:00 AM 7/7/02 12:00 AM 7/8/02 12:00 AM 7/9/02 12:00 AM 7/10/02 12:00 AM -200 Date Figure 3. Airflow manifolds connected to test panel on counterbalance system (Burnett et al [2004]) 30

31 Laboratory Validation (1.6 Lps) 1600 Panel 9 - Weight & Relative Humidity 1400 Weight change Moisture Expert Weight (g) /11/02 12:00 AM 7/13/02 12:00 AM 7/15/02 12:00 AM 7/17/02 12:00 AM 7/19/02 12:00 AM 7/21/02 12:00 AM 7/23/02 12:00 AM 7/25/02 12:00 AM Date Figure 3. Airflow manifolds connected to test panel on counterbalance system (Burnett et al [2004]) 31

32 Field Validation Brick SBPO Ventilated Experimental Overshoot Homosote MC, weight-% Measured Calc /01/02 05/12/02 08/20/02 11/28/02 03/08/03 06/16/03 Date 32

33 Field Validation Vinyl SBPO Ventilated MC, weight-% 15 Measured Calc 10 5 Uncertainty of MC Probe ±4 % Did not Adjust for Initial Conditions 0 02/01/02 05/12/02 08/20/02 11/28/02 03/08/03 06/16/03 Date 33

34 Validation (LAB + FIELD) ME has been validated for Brick & Vinyl Walls Excellent Agreement was found Complex Processes Involved: Liquid Penetration (Incidental Water) Redistribution of Water Ventilation drying Diffusion Transport 34

35 LOADS Question : The greatest UNKNOWNS How much load (water penetration) Does this woodpecker cause? 35

36 LOADS Rely on past data: NOPE Guess : NOPE Evolution in understanding ASHRAE SPC 160 P Yeap! (Systems Approach) 36

37 Moisture Control: Building System Moisture, Energy Production Energy + Air + Moisture HVAC Out 37

38 Application Case IECC Vapor Retarder Recommendation Effort by DOE Building Science Corporation ORNL 38

39 Vapor Retarder Movement In the USA Millions of Homes are in Trouble

40 The Confusion Vapor Guidelines & Codes Vapor Retarders are needed Vapor Retarders are not needed Vapor Retarder are important Vapor Retarder are not important Too many opinions, not enough science 40

41 MOISTURE ENGINEERING 1) Load Based Analysis (Interior & Exterior) 2) Building Envelope System and Subsystems are needed 3) Includes all appropriate physics that describe the transport process 4) Incorporates a saftey factor 41

42 Saftey Factor in Moisture Analysis?? (WOW!!) Factor of safety (FoS), also known as Safety Factor, is a multiplier applied to the calculated maximum load (vapor, rain,water penetration or a combination) to which a component or assembly will be subjected. Thus, by effectively "overengineering" the design by strengthening components or including redundant systems, a Factor of Safety accounts for imperfections in materials, flaws in assembly, material degradation, and uncertainty in load estimates. An alternative way to use the safety factor is to derate the performance (strength) of the material/system to get a "design" strength. Sdesign = Syield / FoS Sdesign = Sproof / FoS 42

43 Margin of Saftey in Moisture Analysis?? (WOW!!) An appropriate factor of safety is chosen based on several considerations. Prime considerations are the accuracy of load and ageing estimates, the consequences of failure, and the cost of overengineering the component to achieve that factor of safety. For example, components whose failure could result in substantial financial loss, serious injury (health consequences or death usually use a safety factor of four or higher (often ten). Non-critical components generally have a safety factor of two. An interesting exception is in the field of Aerospace engineering, where safety factors are kept low (about ) because the costs associated with structural weight are so high. This low safety factor is why aerospace parts and materials are subject to more stringent testing and quality control. Factor of safety of 1 implies no safety at all. Hence some engineers prefer to use a related term, Margin of Safety (MoS) to describe the design parameters. The relation between MoS and FoS is MoS= FoS-1. 43

44 How did we use a Safety Factor Exterior Load Choose Weather Years in a specific Manner Interior Load Investigated three different interior loads Wall & Location Specific Water Penetration in Wall (Dump Water into wall) 44

45 Water penetration: SAFETY FACTOR SPC 160P 1 % water penetration On WRB 1 % water penetration 45

46 Exterior Loads (SF) Employed IEA Annex 24 (10 % Hot & Cold Years) 46

47 Interior Loads (SF) 0.7 Monthly Running Averages SEATTLE BSC: Approach A Relative Humidity (-) Three Loading Conditions Moisture Increment 2 Moisture Increment 4 Moisture Increment Oct. Time (days) 47

48 Interior Loads (SF) 0.7 Monthly Running Averages Chicago BSC: Approach A Relative Humidity (-) Three Loading Conditions.. (low-normal mid average High above average) Moisture Increment 6 Moisture Increment 4 Moisture Increment Oct. Time (days) 48

49 Real Monitored Data SEATTLE: RH (Limits) of 22 Homes had ELEVATED RH DCLU-WSU-(ORNL)

50 Interior Loads (SF) BSC: Approach A Three Loading Conditions MI, g/kg T, C 50

51 We were not Skimpy with Loads 51

52 Simulation Parametric -Part A Absorptive Cladding Non Absorptive Cladding 52

53 Simulation Parametric BSC/Building America Absorptive to Semi Absorptive Cladding 53

54 54

55 55

56 Brick Veneer Interior Coat 8-Perms Seattle CHICAGO Unvented OSB Moisture Content of OSB Sheathing (Interior Location) Moisture Increment 2.5 Moisture Increment 4 Moisture Increment Oct. 1 56

57 Brick Veneer Interior Coat 8-Perms Moisture Content of OSB Sheathing (Interior Location) Seattle CHICAGO Ventilated Exterior Point 1 Point 2 Point 3 Point 4 Interior Point 5 OSB Moisture Increment Oct. 1 57

58 Brick Veneer Interior Coat Asphalt-Kraft Moisture Content of OSB Sheathing (Interior Location) Seattle CHICAGO Ventilated Exterior Point 1 Point 2 Point 3 Point 4 Interior Point 5 OSB 0 Kraft Paper Moisture Increment Oct. 1 58

59 Brick Veneer Asphalt-Kraft Unvented Moisture Content of OSB Sheathing (Interior Location) Exterior Point 1 Point 2 Point 3 Point 4 Interior Point 5 OSB CHICAGO Seattle Kraft Paper Unvented Moisture Increment Oct. 1 59

60 Vinyl Asphalt-Kraft Vented Moisture Content of OSB Sheathing (Interior Location) Seattle Ventilated Point 2 Point 4 Interior Point 5 Vinyl OSB Moisture Increment Oct. 1 60

61 Vinyl None (8-perms) Vented Moisture Content of OSB Sheathing (Interior Location) Seattle Ventilated Point 2 Point 4 Interior Point 5 Vinyl OSB Moisture Increment Oct. 1 61

62 Vapor Retarder Recommendations IECC INTERNATIONAL ENERGY CONSERVATION CODE 62

63 Vapor Retarder Recommendations 63

64 Vapor Retarder Recommendations 64

65 FIELD ANALYSIS VALIDATION Models Need Experimental Data to Validate Their Performance Natural Exposure Test Facility or NET Constructed for this Purpose 65

66 MODEL VALIDATION (Cont) Initial Efforts Focus on Stucco Claddings Potential for Increased Energy Efficiency Overcome Negative PR Regarding Hygric Problems of Energy Efficient Structures 66

67 PANEL S8: EIFS/NO CAVITY INSULATION 67

68 WSU PANELS : THREE COAT PLASTER 68

69 WSU PANELS: ONE COAT STUCCO 69

70 PANELS : FIBER CEMENT SIDING 70

71 Limitation of Experimental Study 1) Reliability of Moisture Content Sensors ( +/- 5 % MC) 2) Reliability of MC sensing zone ( +/- 20 % of MC) 3) No calibration data for OSB, Plywood used 4) Reliability of RH sensors ( +/- 7 %) or higher if exposed to high RH zones 5) Reliability of T ( +/- 1.8 C) Still state-of-the-art 6) Wetting Variation large (Unknown liquid distribution), free water dripping unknown. 71

72 Sensor Location Water Droplets Not accurate sensor Not Possible to describe the exact location for MC measurement 72

73 System Effects Cracks 73

74 1 Year Results Period Oct Oct

75 YEAR 1 75

76 MCc 1 T 1 RHc 4 T 10 MCc 2 T 2 RHc 3 T 9 MCc 3 T 3 MCc 4 T 4 T 11 MCc 5 T 5 MCc 6 T 6 Top Plate Inside Sheathing Outside Sheathing / Out Sheathing Inside Stud Bottom Plate 76

77 Results: Exterior & Interior Loads 77

78 Wall 1: Stucco, Poly, Unvented, OSB 1-top pl, 2-OSB, 3-OSB_out, 4-OSB, 5-Stud, 6-bot pl 78

79 Wall 3: Stucco, Poly, Vented, OSB 1-top pl, 2-OSB, 3-OSB_out, 4-OSB, 5-Stud, 6-bot pl 79

80 Wall 5: Stucco,Kraft,Unvented, Plywood, R-11(2x4), Oil Paint 1-top pl, 2-Ply, 3-Ply_out, 4-Ply, 5-Stud, 6-bot pl 80

81 Wall 7: Stucco,No VR,Unvented,OSB 1-top pl, 2-OSB, 3-OSB_out, 4-OSB, 5-Stud, 6-bot pl 81

82 Wall 4: Stucco, Poly, Ventilated, OSB 1-top pl, 2-OSB, 3-OSB_out, 4-OSB, 5-Stud, 6-bot pl 82

83 More investigations Two classes of Claddings Unvented Stucco, Absorptive Claddings Ventilated Stucco, Board Claddings 83

84 RH interior Notice- High RH at inner surface of insulation 55 % to 85 % (5 months/year May to Oct) 84

85 Conclusion- 1 Year Results Ventilation increases drying potential by a factor of 3 (drying rate using stucco loading in Seattle) Venting increases drying performance by 33 % Foam insulation keeps wall warm and increases drying performance Seattle requires an interior vapor retarder (at least kraft coating unless cladding ventilated), Membrain outperforms poly and no vapor retarder Old constructions also prone to moisture problems if use similar stucco No Vapor Retarder (60-perms or 10 as thought of when built) 85

86 Conclusion- 1 Year Results Membrain outperforms poly and no vapor retarder No Vapor Retarder (60-perms or 10 as thought of when built) Field and model must be used together to develop code changes Do not use ONLY models but field and models to produce code changes that are needed 86

87 Future WUFI-plus Whole Simulation Model 87

88 What is next? Climate Data Construction Data Material Data Whole Building Simulation Hygrothermal Envelope Simulation (WUFI) Damage & Aging Models Energy Consumption, Comfort Service Performance (biological, chemical mechanical resistance) Durability, Rentability 88

89 WUFI-Family 2006 Hygrothermal Building Simulation Hygrothermal Envelope Simulation Postprocessing Models (Result Evaluation) WUFI+ 1.0 WUFI-1D (ORNL/IBP, Pro) 3.2 WUFI-2D 2.1 WUFI-Bio 1.0 WUFI-Star WUFI-CFD 89

90 COMBINING THERMAL BUILDING SIMULATION AND HYGROTHERMAL ENVELOPE CALCULATION Envelope Heat w φ = div ) φ t ( D grad( φ δ grad( φ )) φ + p p sat 90

91 91 COMBINING THERMAL BUILDING SIMULATION AND HYGROTHERMAL ENVELOPE CALCULATION Whole Building Heat ( ) vent i a l i Sol i j j j j i Q c V n Q Q A dt d V c & & & = θ θ ρ θ θ α θ ρ ) ( Moisture ( ) Vent IMP i a j w j j i W W c c V n g A dt dc V =

92 MODEL VALIDATION foil faced test room reference room 92

93 MODEL VALIDATION Relative Humidity [%] Measurement Simulation Time [h] 93

94 Future Modeling It was always considered: indoor humidity is of small importance for a successful design because temperature is easier to sense, quantify and comprehend. Indoor relative humidity (RH) is important and has significantly impact. New model for : Transient behavior for the whole building, its indoor climate AND the envelope With the model it is possible to make more and more accurate predictions of the indoor humidity variations 94

95 Conclusions Models are as experienced as the operator who uses them (Training is necessary) There are many limitations to models There are even more limitations for testing Field and Modeling should be used for CODES In the near future WUFI-PLUS will assist in Whole Building Design Work is proceeding with a 3-D CFD WUFI model with CAD interfaces 95