Science and research
Theoretical research
Theoretical research takes place simultaneously with experimental measurements of physical and mechanical properties of the engineering structure of the timber construction. The aim of the theoretical research is to verify the accuracy of applied calculation methods to define e.g. a temperature field inside an engineering structure in comparison with real results from experimental measurements.The purpose of the research is to verify the suitability of the applied numerical simulation method to predict the thermal behaviour of light engineering structures based on timber, which has a significant impact on temperature contentment inside a building.
An implicit numerical method was selected to analyse the thermal behaviour of the circumferential construction theoretically. This numerical implicit method of finite elements is more suitable for non-stationary multi-measurement thermal calculations than the analytical method. The method is based on discretization of the space (resp. the time). Numerical methods allow us to obtain a solution for the thermal (or temperature) problem with a final number of discrete places (junctions) in a selected network covering either the whole area or just a selected part. Programmes ANSYS 12 and AREA 2011 were used for the numerical simulation calculation.
- The numerical simulation of the temperature behaviour of the circumferential construction of the timber building was performed for 3 selected construction details of the timber building which differ in their material composition and orientation towards the four cardinal directions:
- detail 1 (circumferential wall , the south facade)
- detail 2 (the corner of circumferential walls, the north-west facade)
- detail 3 (roof construction)
All details were evaluated for both the summer and winter seasons of the year. A theoretical analysis of all details was performed for real boundary conditions in the summer and in the winter which were obtained through experimental measurements.
The whole process of the preparation and calculation of the simulation model in ANSYS programme can be divided into three basic stages:
1. Pre-processing – preparation of input data. The solved models of circumferential structures were created using type PLANE55 surface finite elements (PLANE55 Numerical Model circumferential wall , PLANESS55 Numerical Model roof (detail 3)). Each material was defined using thermo-technical properties: volumetric weight, specific heat capacity, coefficient of thermal conductivity. Particular attributes were allocated to individual surfaces under the network creation; the whole element was divided with a network and junctions were created.
PLANE55 Numerical Model circumferential wall (detail 1)
PLANESS55 Numerical Model roof (detail 3)
2. Processing – calculation (numerical core). The analysis of the calculation was solved as a two-dimensional stationary and non-stationary task. The non-stationary task requires an input of the time and the time step. The minimum time step of 100s and the maximum time step, identical with the time step of the measurements (3,600 s), were used for the calculation. The total length of the evaluated time segment was set at 24 hours.
The initial condition of the calculation, which describes a distribution of temperature within a body at the beginning of an action with time t0, was calculated with a reference temperature θ = 30°C for the summer season. This temperature was verified with simulation calculations so that the initial thermal condition of the structure could correspond with real measured temperatures in time t0. For the winter season, the initial condition for the temperature distribution within a construction was first set through a simulation calculation of a thermal field under stationary conditions (for detail 1: θse = -6,5°C, θsi = 18,8°C, for detail 2: θse = -5,0°C, θsi = 17,1°C) and this state was further used as an initial condition for a nonstationary calculation. The initial condition of a temperature field distribution was verified through simulation calculations so that the initial thermal condition of the structure in time t0 could correspond with real measured temperatures.
The boundary conditions for the non-stationary calculation for the summer season were defined for the external side of the engineering structure using the Dirichlet first-type condition – the external surface temperatures of the structure were used from the real measured hourly rates on 8. 8. 2013 (from 1:00 o’clock until 24:00 o’clock). For the internal side of the engineering structure, the Dirichlet first-type condition was set – the internal surface temperature of the structure θsim = 32,1°C, which was considered to be constant over the whole time segment of 24 hours for detail 1 and the constant surface temperature θsim = 33,2°C was considered constant for detail 2.
3. Post-processing – data processing. Figure 4.3. shows the course of temperature field in the construction details at 14:00 on 8. 8.
Temperature field in the circumferential wall (detail 1)
Temperature field in the corner (detail 2)
Temperature field in the roof (detail 3)
Verifying results from the theoretical research and experimental measurements
Tables 4.1 and 4.2 compare results of numerical calculations of the temperature field and experimental measurements in detail 1(in places of sensors) on the selected summer day of 8.8. and on the selected winter day of 26. – 27. 1. Both days were selected with respect to outside temperatures whose maximum and minimum readings were approaching standard temperature readings according to.
Tab. 4.1 Results from the experimental measurements and numerical calculations of the temperature field in detail 1 in the summer season (on 8.8.2013)
| time | no. of the temperature sensor position | temperature | [°C] | |||
| [hod] | 1 | 2 | 3 | 4 | 5 | |
| 1:00 | Experim. | 32,0 | 31,1 | 30,1 | 29,7 | 21,7 |
| Ansys | 32,1 | 30,1 | 30,0 | 29,9 | 21,7 | |
| 5:00 | Experim, | 31,7 | 31,1 | 30,1 | 28,8 | 19,8 |
| Ansys | 32,1 | 30,5 | 30,0 | 29,8 | 19,8 | |
| 9:00 | Experim, | 31,6 | 31,1 | 29,8 | 28,0 | 27,1 |
| Ansys | 32,1 | 30,8 | 30,1 | 29,3 | 27,1 | |
| 13:00 | Experim, | 31,9 | 31,0 | 29,5 | 27,7 | 41,6 |
| Ansys | 32,1 | 31,1 | 30,1 | 29,9 | 41,6 | |
| 15:00 | Experim, | 32,3 | 31,0 | 29,5 | 28,3 | 43,8 |
| Ansys | 32,1 | 31,1 | 30,1 | 29,1 | 43,8 | |
| 19:00 | Experim, | 32,9 | 31,2 | 29,9 | 29,8 | 34,4 |
| Ansys | 32,1 | 31,3 | 30,2 | 29,9 | 34,4 | |
| 23:00 | Experim, | 32,7 | 31,5 | 30,5 | 30,5 | 24,8 |
| Ansys | 32,1 | 31,4 | 30,5 | 30,6 | 24,8 | |
The above results show identical course of temperatures measured and calculated in individual positions of temperature sensors. The research results proved that dynamic simulation methods are suitable for temperature analysis of engineering structures.
Tab. 4.2 Results from the experimental measurements and numerical calculations of the temperature field in detail 1 in the winter season (on 26. – 27. 1.)
| time | no. of the temperature sensor position | temperatures | [°C] | |||
| [hod] | 1 | 2 | 3 | 4 | 5 | |
| 6:00 | Experim. | 18,7 | 15,6 | 8,4 | 1,2 | -7,3 |
| Ansys | 18,8 | 15,5 | 8,3 | 1,1 | -7,3 | |
| 10:00 | Experim. | 18,8 | 15,8 | 8,4 | 1,1 | -7 |
| Ansys | 18,8 | 15,7 | 8,8 | 1,0 | -7 | |
| 14:00 | Experim. | 18,8 | 15,6 | 8,4 | 1,0 | -3,2 |
| Ansys | 18,8 | 15,5 | 8,5 | 0,9 | -3,2 | |
| 18:00 | Experim. | 18,8 | 15,6 | 8,3 | 1,1 | -7,6 |
| Ansys | 18,8 | 15,5 | 8,4 | 1,0 | -7,6 | |
| 22:00 | Experim. | 18,8 | 15,6 | 8,4 | 1,2 | -12,9 |
| Ansys | 18,8 | 15,5 | 8,3 | 1,1 | -12,9 | |
| 2:00 | Experim. | 18,7 | 15,6 | 8,4 | 0,9 | -13 |
| Ansys | 18,8 | 15,5 | 8,3 | 1,0 | -13 | |
| 5:00 | Experim. | 18,7 | 15,6 | 8,2 | 0,4 | -10,1 |
| Ansys | 18,8 | 15,7 | 8,5 | 0,2 | -10,1 | |
Table 4.3. compares results from the numerical calculation for various boundary conditions (standard stationary conditions – AREA programme and measured non-stationary conditions – ANSYS programme in construction detail 1 and experimental measurements. The calculation was performed for the winter season because standard calculation methods issue from the solutions during stationary boundary conditions and it was possible to compare them. For construction detail 1, the calculation was performed in two transverse profiles of walls – in the axis between two wooden columns and in the place of a wooden column so that the effect of a thermal bridge on the temperature field in the structure could be evaluated.
Tab. 4.3 Comparison of results from the experimental measurement and the numerical calculation
| Temperature in the axis between the colummns [°C] | |||||
| No. of the temperature sensor position | 1 | 2 | 3 | 4 | 5 |
| Experimental measurement | -8,9 | -1,9 | 7,2 | 15,6 | 19,1 |
| Calculation in AREA programme | -11,6 | -5,9 | 5,4 | 15,4 | 19,7 |
| Calculation in ANSYS programme | -8,0 | -2,0 | 7,0 | 15,6 | 19,6 |
| Temperature in the place of the column [°C] | |||||
| Experimental measurement | -8,8 | -0,2 | 8,4 | 15,2 | 18,7 |
| Calculation in AREA progremme | -11,6 | -5,0 | 5,2 | 15,8 | 19,2 |
| Calculation in ANSYS programme | -8,0 | -3,4 | 6,0 | 15,9 | 19,2 |
The results prove that the calculation processes considering the measured non-stationary boundary conditions approach the measured quantities more than the calculations considering standard non-stationary boundary conditions.
The above listed readings and temperature outputs from AREA and ANSYS programmes clearly show that the impact of the thermal bridge (the bearing wooden column) in the circumferential wall demonstrates itself in an increased transmission of heat in a particular area. The difference between temperatures in individual parts of the engineering structure oscillates between 0,1 and 1,7°C for the measured values. It is problematic to compare results from the measurements and individual calculation processes, due to difficulties in mutual comparisons of boundary conditions, still we can see that the calculation processes when considering the non-stationary transfer of heat get closer to the measured quantities than the stationary calculations.