热设计网

EFD介绍及EFD之实例验证

FloEFD


EFD介绍及EFD之实例验证

势流科技 资深CAE工程师 杻家庆(Jason Niu)
 
热设计 https://www.resheji.com
Agenda
 Introduction of EFD
 Key technologies of EFD
 Application in EFD
 EFD之实例验证
 
Introduction of EFD
What is EFD??
 EFD指的是Engineering Fluid Dynamics
 EFD是一套简单易学的热流分析软件
 对于工程师在CFD分析背景的要求较低
 直接将CAD模型导入CFD的概念,使设计工程师可以在设计前期找到最佳化的模型
 EFD是一套专为改善研发设计前端的热流分析软件

Typical Development Costs

EFD Product Family
 EFD. Lab: 使用solidworks接口
 EFD. Pro:使用Pro/ENGINEER接口
 EFD.V5: 使用CATIA V5接口
 EFD. Flexx:适合使用不同CAD软件的user,可直接利用floating的方式到不同的CAD接口

Key Technologies of EFD
 DC3 –直接将CAD的模型导入CFD的概念
 RAM –矩形网格自动生成
 MWF –改良边界层的计算函数
 LTTM –层流渐变到紊流的精确计算模式
 ACC –自动收敛控制
 DVA –参数化设计分析
 EUI –工程师使用平台

Direct CAD to CFD Foundation
 直接使用PRO/E、solidworks、CATIA建立模型
 自动区分固体与流体空间
 自动检查内流场与外流场区域,无须在CAD中建立流场区域

Rectangular Adaptive Meshing
 自动对流体与固体区域划分网格
 自动针对物理特性与外型特征进行网格加密
 No “black box” approach亦可利用参数控制网格的生成Base mesh Refine mesh

Modified Wall Function
 接近边界层的网格采用partial cells技术定义
 物理修正流动与热传现象的边界层效应的模拟EFD simulation Experiment

Laminar–Transitional–Turbulent Modelling
 采用改良边界层的计算函数直接仿真层流与紊流的现象
 不须指定流体的特征,即可在同一个模型仿真出由层流 渐变流 紊流的现象

Automatic Convergence Control
 使用者可以针对模型选择某面、某对象、计算域的任何变量作收敛控制
 使用者可以针对具代表性的方程式作收敛控制
 不须作任何数值方法的控制即可完成收敛

Design Variant Analysis
 采用” what-if ”的设计概念,直接利用参数对模型进行优化设计
 优化模型可直接被CAD采用

Engineering User Interface
 Easy-to-use:采用CAD平台直接进行CAE模拟
 Easy-to-learn:以工程师角度设计,使用工程师语言,让工程师容易学习
 深入的结果解析与图形显示功能
 直接产生MS-Office格式的报告

These features are available for the Electronics module users only.

Electronic Database
 增加电子产品相关数据库
 包含material、Fan curve、Heat pipe、PCB…等

Perforated Plate
 可以选定对象的面设定其为2D Resistance。
 设定其free area ratio与hole shape等

Two-Resistor Component
 利用选定Junction与case对象,以及设定junction 的性质,即可模拟two-resistor component。
Two-Resistor model Normal model

Electrical Condition
 利用设定current、voltage或contact resistance,模拟焦耳热功问题。

Heat Pipe
 仅须选定heat pipe part,定义加热端与散热端,并设定有效热阻值(effective thermal resistance),即可模拟热管。

Printed Circuit Board
 PCB(Printed Circuit Board)模块可以直接设定PCB各层的铜覆盖率,更方便且精确模
拟PCB散热情形。


Application in EFD
Simulation of chip thermal state
Thermal model of a chip
kx=ky=Solder and Air 25*25*0.8 mm3 y=0.034; kz=11.2 W/(m*K)
Substrate 25*25*1.14 mm3 kx=ky=9.9; kz=2.95 W/(m*K)
Bump and Underfill 8*7.3*0.07 mm3 kz=5 W/(m*K)
Die 8*7.3*0.86 mm3 Silicon

Surface heat source: 15 W
In considered tasks ambient temperature was set as 20 ??, gravitation and radiation were taken into account


Chip with heat sink and interconnection simulation
PCB has three layers. Thickness of each layer – 0.5 mm.
Interconnection thickness – 50 microns.
Heat sink [°C] 99
Pcb 1 layer [°C] 44
Bump and Underfill [°C] 103
Solder and Air [°C] 87
Substrate [°C] 88
Die [°C] 103
Number of mesh cells 252,724 Computational time ~ 2.1 h

Thermal Management for Projectors

Natural Convection with Radiation
Heat model of headlight
Screen
Headlight
Lamp with power of 30 W
-The heat radiation from the light source;
-The rays penetration through the transparent
bodies;
-The rays reflection from the reflective surfaces;
- Radiation heat exchange between bodies;
- Natural convection process;
- Heat conductivity in solid bodies.

Model of the headlight
Aluminum Headlight
housing
Bulb and
headlight glass
Volume inside the
lamp
Filament
Parabolic reflective surface
  The filament has temperature of 2500 K
  The screen is modeled by the low
conductivity material with emissivity of
0.8
 
Temperature at the screen
  Diameter of the hot spot on the
screen corresponds to the
reflector diameter.
  There is the temperature tail
which corresponds to the natural
convection flow near the screen

Temperature and convection flow field
 
Passenger compartment ventilation
Back face
Pressure 1 bar
Inlet
Air velocity 0.5 m/s

Passenger compartment heat transfer task
Housing material – Stainless Steel
Passenger compartment-Result
Satellite exposure to sun radiation
Trajectory
Re= 6300 km
Ro=48000 km
Heat source 1.5 kW
Satellite material – Aluminum
Emissivity coefficient = 0.7

国防工业
CALCULATION OF F16 AT M=0.6 AND 5o (H~500 m)

建筑业
 建筑流场与太阳辐射分析 机械工业
 压缩机模拟
 控制阀模拟

机械工业
 Water mixer
 Heat exchanger
EFD 模拟与实验比较

 感谢不具名的客户提供
Agenda
  前言
  实际案例测试
 Case 1
 Case 2
 Case 3 & 4
(Thermal performance prediction of design changed)
 
前言
  热流数值仿真于近年来在电子3C产业之研发中使用越发频繁,如何快速、准确的预估是研发(R&D)工程师首重考虑之部分。而目前市面上各软件商无不发展适合工程师使用之热流数值仿真软件。
  热流数值仿真软件之选择,首重于研发体系中热流模拟阶段所占之重要性,以及即将使用此软件之工程师共同试用后之评价。
  最后先透过两个旧案进行数值分析与实验之比较,了解软件之功能与实际应用之状况。另外再透过一个开发中之案件协助工程进行判断依客户要求减轻重量而减少鳍片数量后对效能之影响,提供快速的结果比较。另外一探讨不同阶段模型对于趋势预测之影响。
产品实际案例Case 1
PQ-fan
分析的设定如下:
1.发热瓦数62W
2.环温42℃
3.cpu与sink间无建构grease
4.风扇使用PQ曲线分析所需时间:
1.建构与设定分析模型:
20min (不含3D pro/e檔建构)
2.求解: 约2hr
3.后处理: 20min
 
Case 1 simulation result
5% error
产品实际案例Case 2
PQ-fan
分析的设定如下:
1.发热瓦数115W
2.环温35℃
3.cpu与sink间无建构grease
4.风扇使用PQ曲线分析所需时间:
1.建构与设定分析模型: 40min(不含3D pro/e檔建构)
2.求解: 约3.5hr
3.后处理: 25min
 
Case 2 simulation result
4% error
△T ofheat-pipe isabout6 oC
Thermal performance prediction of design
changed
 
Agenda
  Purpose
  Thermal Experiment
  Simulation result
  Discussion
  The influence of fan duct
Purpose
  Due to the weight of cooler, reduce the fin number to reduce the weight.
  Use the thermal fluid analysis tool EFD.PRO to evaluate the temperature difference before mockup .
  Software validation test also.
 
Model drawing
Origin: 112 fins Modified: 84 fins
Thermal Experiment
  Thermal test platform and test sample
LGA775 TTV 8025 axialfan Heatsink case 3 & case 4
Assem bly sam ple
Fan duct
Fan cover
 
Simulation model
  Ta=35 oC
  CPU Power =65W
  Fan =P-Q fan with 4500RPM
Fan (flow) direction is bottom to top Simulation
domain CPU
Simulation result of Model 1 (Origin) Surface plot Temperature section plot Vector plot Temperature section plot
 
Simulation result of Model 2
(fin number reduced)
Surface plot Temperature section plot Vector plot Temperature section plot Temperature comparison
  Section plot of temperature (fix scale)
Temperature difference is about 5oC between two model.
 
Discussion - Data comparison
  The difference between experiment and simulation in absolute value of temperature is about 5oC.
  The thermal resistance difference in two heat sink are 1.80
and 1.73 of experiment and simulation. The trend of simulation is cohere with experimental data.
Improvement of model constriction
  Add a Fan duct to close the real model
 
Flow pattern
  We can see that the flow pattern is quite different. The hot air will not return into heat sink, the result should be more close to real test.
Simulation result
Surface plot Temperature section plot
Vector plot Temperature section plot
 
Temperature result
  Temperature of the simulation with fan duct is close to experimental data. Temperature of the simulation with fan duct is a little higher than experimental. When add a fan cover, the temperature should lower than experimental.
This result is more reasonable.
Conclusion
  配合实验数据的验证,证明EFD在进行thermal module的模拟时,具有一定的准确性;加上EFD可以直接将CAD建构的模型进行散热分析,因此对于近来几何外型有愈来愈复杂的thermal module而言,不需要做太多的外型简化,即可仿真,对于增加产品开发的效益有实质的帮助。

 FloEFD资料下载:  EFD介绍及EFD之实例验证.pdf  

标签: 点击: 评论:

留言与评论(共有 0 条评论)
   
验证码: