TEMTIS Seminar in Horsens September 11 2008 The

TEMTIS Seminar in Horsens September 11, 2008 The effect of Eurocode 5 on timber structure design in Norway Kolbein Bell and Kjell Arne Malo NTNU, Norway September 2008 1 Department of structural engineering

Objective To compare some important passages in the current Norwegian timber code NS 3470 -1 (5 th ed. July 1999) with the current version of Eurocode 5 (EC 5) EN 1995 -1 -1 including EN 1995 -1 -1: 2004/A 1 and to point out some problem areas September 2008 2 Department of structural engineering

Important NOTES • NS 3470 -1 was drafted on the basis of the first draft of EC 5 – the ”philosophy” is therefore much the same • A corrigendum to NS 3470 -1 which will bring the Norwegian code closer to EC 5 is about to be made an official part of NS 3470 -1 – this in order to ”soften” the effects of the transition to EC 5. This corrigendum is disregarded in this presentation September 2008 3 Department of structural engineering

September 2008 4 Department of structural engineering

Load duration NS 3470 -1 P perm. A long t. 1 0, 6 0, 7 0, 8 0, 9 1, 0 1, 1 2 0, 6 0, 7 0, 8 0, 9 1, 0 1, 1 3 0, 5 0, 6 0, 55 0, 65 0, 7 0, 8 0, 9 Service class EC 5 September 2008 B C med. t. short t. 5 I inst. Department of structural engineering

NS 3470 -1 (from 1, 1 to 1, 32) 1, 0 or 1, 1 or 1, 2 for solid timber EC 5 September 2008 for glulam 6 Department of structural engineering

Stability - NS 3470 -1 Combined axial compression and bending: Bending: September 2008 7 Department of structural engineering

Stability - EC 5 Combined axial compression and bending: Bending: September 2008 8 Department of structural engineering

The factor km Rectangle: (bending about two axes) this factor is not present in NS 3470 -1 September 2008 9 Department of structural engineering

Comparison – simple column – P only p=0 NS 3470 -1 GL 36 c C/2 Pu = 188, 2 k. N A/3 Pu = 125, 5 k. N EC 5 September 2008 10 C/2 Pu = 178, 3 k. N A/3 Pu = 111, 5 k. N Department of structural engineering

Comparison – simple column – P & p p = const. = 3, 0 k. N/m NS 3470 -1 GL 36 c C/2 Pu = 120 k. N A/3 Pu = 60, 6 k. N EC 5 September 2008 11 C/2 Pu = 139, 3 k. N A/3 Pu = 72, 5 k. N Department of structural engineering

Compression perpendicular to grain The characteristic strength in NS 3470 -1 is more than twice that of EC 5 (for all strength classes), but NS 3470 -1 uses the actual contact area in the calculation of The formula and the values of factor are not all that different (providing the requirements of “supplement” A 1 are used). More about this later. September 2008 12 Department of structural engineering

Shear The introduction of an effective width by EC 5 (”supplement” A 1), may, dependig on the national choice for the value of , have a significant influence on shear design. September 2008 13 Department of structural engineering

Special glulam components For curved and pitched cambered beams EC 5 has the following formula for combined shear and tension perpendicular to grain: (6. 53) For glulam: September 2008 where 14 Department of structural engineering

NS 3470 -1 has no such formula, nor does it have the two factors and in particular is troublesome; it is both difficult to determine and it seems to have a very significant (adverse) effect for large components). Formula (6. 53) will have a detrimental effect on typical Norwegian arch bridge designs. September 2008 15 Department of structural engineering

Example: Glulam arch bridge - loading September 2008 16 Department of structural engineering

Bending moment (M) and shear force (V) M problem V September 2008 17 Department of structural engineering

Prior to formula (6. 53) EC 5 had the formulation: …. shall be satisfied. In the recently approved ”supplement” A 1 this has been changed to: …. should be satisfied. Exactly how should the designer interpret this? September 2008 18 Department of structural engineering

Connections EC 5 has similar, but more complex formulas than NS 3470 -1. The most noticeable differences are: • in NS 3470 -1 the first 6 fasteners are effective • NS 3470 -1 does not take account of the rope effect • NS 3470 -1 does not recognize block or plug shear Our experiences so far suggest minor differences, but no systematic bias either way. September 2008 19 Department of structural engineering

More about compression perpendicular to grain Comparison EC 5(A 1) vs NS 3470 -1 September 2008 20 Department of structural engineering

geometric parameters: design parameters (ULS): September 2008 21 Department of structural engineering

Strength ultimate limit state (ULS) September 2008 22 Department of structural engineering

Design: NS 3470 -1 & EC 5/A 1(2004) NS 3470 -1 EC 5/A 1(2004) Solid Wood ? ? ? Parameters area 0. 5 – 1. 8 1. 0 - 1. 75 C 14 4. 3 2. 0 C 18 4. 8 2. 2 C 24 5. 3 2. 5 C 30 5. 7 2. 7 1. 1 (1. 21) 1. 3 material factor 0. 9 0. 8 SC 2, medium term September 2008 23 factor strength Department of structural engineering

Solid wood, C 24: September 2008 24 Department of structural engineering

EC 5/A 1(2004) CASE VALUE General 1, 0 Special a) cont. support 1, 25 1, 50 Cxx GL Special b) discr. support 1, 5 1, 75 Cxx GL September 2008 25 MATERIAL REQUIRED Department of structural engineering

NS 3470 -1 • For • is • where: • while for • is September 2008 26 Department of structural engineering

Solid wood, Case 1: • C 24 • Continous support • Vertical column on end of beam September 2008 27 Department of structural engineering

Capacity pr unit width [ N/mm ] EC 5 / A 1(2004) September 2008 28 Department of structural engineering

Capacity pr unit width [ N/mm ] NS 3470 -1 September 2008 29 Department of structural engineering

Strength ratio: EC 5(A 1) / NS 3470 -1 September 2008 30 Department of structural engineering

Solid wood, Case 2: • C 24 • Continous support • Vertical column on continuous beam September 2008 31 Department of structural engineering

Capacity pr unit width [ N/mm ] EC 5 (A 1) September 2008 32 Department of structural engineering

Capacity pr unit width [ N/mm ] NS 3470 -1 September 2008 33 Department of structural engineering

Ratio EC 5(A 1) / NS 3470 -1 Colums internal on cont. sup. beam September 2008 34 Department of structural engineering

Solid wood (C 24), Case 3: • Vertical load transfer through beam section (h > a) • Beam continuous • EC 5: (? ) September 2008 35 Department of structural engineering

EC 5 (A 1) NS 3470 -1 [ N/mm ] September 2008 36 Department of structural engineering

Ratio EC 5(A 1) / NS 3470 -1 (C 24) Colums internal on beam September 2008 37 Department of structural engineering

Solid wood (C 24), Case 4: • Vertical load transfer through beam section at the beam end • EC 5? (cover this case? ) September 2008 38 Department of structural engineering

Ratio EC 5 (A 1) / NS 3470 -1 Columns at beam end September 2008 39 Department of structural engineering

EC 5 (A 1): Better capacity for eccentric load transfer (C 24) September 2008 40 Department of structural engineering

GLULAM Examples: GL 32 c September 2008 41 Department of structural engineering

Ratio EC 5 / NS 3470 -1 GL 32 c ”Column at beam end, cont. sup. ” September 2008 42 Department of structural engineering

Ratio EC 5 / NS 3470 -1 GL 32 c ”Column internal on beam, cont. sup. ” September 2008 43 Department of structural engineering

Ratio EC 5 / NS 3470 -1 GL 32 c ”Column connections at beam end” September 2008 44 Department of structural engineering

Ratio EC 5 / NS 3470 -1 GL 32 c ”Column connections internal on beam” September 2008 45 Department of structural engineering

Ratio EC 5 / NS 3 470 -1 GL 32 c ”Columns eccentric internal on beam, discrete. sup. ” September 2008 46 Department of structural engineering

Compression perpendicular to grain Concluding remarks • EC 5 compared to NS 3470 -1 gives for: – Solid wood: • roughly only 2/3 of the capacities • small capacities for vertical load transfer through horisontal beams • enhanced capacities for colums at beam ends • highest capacities for small contact length due to the effektiv length concept – GLULAM: • Overall similiar to solid wood, but the difference is smaller • Higher capacities for small contact length (< 30 mm) • Smaller capacity for vertical load transfer through continuous beams September 2008 47 Department of structural engineering

Summary – ultimate limit state Generally speaking, ULS-design of timber structures by EC 5 will result in somewhat more conservative designs than NS 3470 -1. We are talking about 5 to 25 %, most of which is caused by and. In some special cases the effect can be much higher. Our experiences over the past couple of decades do not seem to warrant this ”extra safety”. September 2008 48 Department of structural engineering

Some additional problems Serviceability limit state design, as specified by EC 5, is rather complex and error prone. EC 5 is not particularly well suited for more accurate, nonlinear static analyses as basis for design (nor is NS 3470 -1). Major issues are: - stiffness parameters (E and G) - shape and size of geometric imperfections - modelling of joints - failure criteria September 2008 49 Department of structural engineering

Consider • E (G) as “computational” parameter(s), accounting for all factors influencing the stiffness of the structural members • ultimate load design • solid timber and glulam of softwood September 2008 50 Department of structural engineering

EC 5 - some definitions Mean value: Characteristic (fifth percentile) value: Final mean value: Design value: - partial factor for a material property - factor for quasi-permanent value of an action - factor for the evaluation of creep deformation September 2008 51 Department of structural engineering

NOTE: EC 5 specifies: for solid timber for glulam September 2008 52 Department of structural engineering

EC 5 states that the analysis of a structure shall be carried out using the following values for stiffness properties: 1 st order linear elastic analysis of structure whose distribution of internal forces is insensitive to stiffness distribution 1 st order linear elastic analysis of structure whose distribution of internal forces is affected by the stiffness distribution 2 nd order linear elastic analysis of structure September 2008 53 Department of structural engineering

Case - Simply supported column Glulam GL 36 c strong axis September 2008 54 Department of structural engineering

EC 5 - nonlinear approach Geometric imperfection in the shape of the 1 st buckling mode L e = L / 400 e (EC 5 suggests L / 500) For E = 12000 Mpa PE = 304, 7 k. N for L = 4 m PE = 135, 4 k. N for L = 6 m September 2008 55 Department of structural engineering

Capacity (Pult) for long-term load in service class 3 q=0 L=4 m e = 10 mm medium slenderness: Linear analysis: Pult = 122, 7 k. N Nonlinear analysis ( kc, y = kc, z = 1, 0 ): September 2008 56 Department of structural engineering

Capacity (Pult) for long-term load in service class 3 q=0 L=6 m e = 15 mm large slenderness: Linear analysis: Pult = 56, 5 k. N Nonlinear analysis ( kc, y = kc, z = 1, 0 ): September 2008 57 Department of structural engineering

Capacity (Pult) for long-term load in service class 3 q = 1 k. N/m L=6 m e = 15 mm large slenderness: Linear analysis: Pult = 43, 0 k. N Nonlinear analysis ( kc, y = kc, z = 1, 0 ): September 2008 58 Department of structural engineering

Long-term load in service class 3 q=0 L=6 m September 2008 e = 15 mm 59 Department of structural engineering

q=0 L=6 m e = 15 mm S/1 L/3 S/1 September 2008 60 Department of structural engineering

This simple example seems to leave a few questions: - size of imperfection? - what is the appropriate stiffness? and are factors , and parameter really independent of load duration and service class, or should they be functions of in some ways? September 2008 61 Department of structural engineering

Thank you September 2008 62 Department of structural engineering