Basic Observable Measurement Functions
[Tools]

Functions
float	getHarmonicPotentialEnergy (float dEpot, FLOAT_ARRAY_TYPE dPos, float sum, int cnn, size_t &pitch, int &N, int &numNN, float &k, float &boxSize, dim3 &dimGrid, dim3 &dimBlock, int &maxThreads, int &maxBlocks)
	Calculates total harmonic potential energy.
float	getKineticEnergy (FLOAT_ARRAY_TYPE dVel, int &N, float sumSq, int &maxThreads, int &maxBlocks)
	Calculates total kinetic energy.
float	getLJCutPotentialEnergy (FLOAT_ARRAY_TYPE dPos, float dEpot, float *sum, int &N, float &boxSize, float &rCutSq, dim3 &dimGrid, dim3 &dimBlock, int &maxThreads, int &maxBlocks)
	Calculates the total Lennard-Jones energy of the system.
__global__ void	gpuGetHarmonicPotentialEnergyKernel (float dEpot, FLOAT_ARRAY_TYPE dPos, int *cnn, size_t pitch, int N, int numNN, float k, float boxSize)
	Kernel for harmonic potential.
__global__ void	gpuLJCutPotentialEnergyKernel (FLOAT_ARRAY_TYPE dPos, float dEpot, int N, float boxSize, float rCutSq)
	Calculates Lennard-Jones potential within a certrain cutoff.

Detailed Description

This module contains basic measurements of energies

Function Documentation

float getHarmonicPotentialEnergy	(	float *	dEpot,
		FLOAT_ARRAY_TYPE *	dPos,
		float *	sum,
		int *	cnn,
		size_t &	pitch,
		int &	N,
		int &	numNN,
		float &	k,
		float &	boxSize,
		dim3 &	dimGrid,
		dim3 &	dimBlock,
		int &	maxThreads,
		int &	maxBlocks
	)

Calculates total harmonic potential energy $E_{harm} = \frac{1}{2} \sum_{\left<i,j\right>} kr_{i,j}$

Parameters:

`[out]`	dEpot	Array on the device to hold the total potential energy for each particle
`[in]`	dPos	Particle position array
`[in]`	sum	Summation variable on the device used for intermediate data during reduction, content will be overwritten
`[in]`	cnn	Array holding the indices of harmonic linked particles
`[in]`	pitch	Memory alignement pitch on the device
`[in]`	N	Number of particle
`[in]`	numNN	maximum number of connected neighbors
`[in]`	k	Spring constant
`[in]`	boxSize	Edge length of the cubic simulation box
`[in]`	dimGrid	Grid dimension
`[in]`	dimBlock	Block dimension
`[in]`	maxThreads	Maximum number of allowed threads
`[in]`	maxBlocks	Maximum number of allowed blocks

Returns:: Returns total harmonic potential energy of the system.

Definition at line 269 of file gpu_tools.cu.

References checkCudaError(), and sumReduction().

float getKineticEnergy	(	FLOAT_ARRAY_TYPE *	dVel,
		int &	N,
		float *	sumSq,
		int &	maxThreads,
		int &	maxBlocks
	)

Calculates total kinetic energy $E_{kin} = \frac{1}{2} \sum_{i=1}^N m_i \vec{v_i}^2$ , where $ m_i=1$ due to Lennard-Jones units.

Parameters:

`[in]`	dVel	Velocity array
`[in]`	N	Number of particles
`[in]`	sumSq	Helper variable, content will be overwritten
`[in]`	maxThreads	Maximum number of allowed threads
`[in]`	maxBlocks	Maximum number of allowed blocks

Returns:: Returns total kinetic energy of the system.

Definition at line 255 of file gpu_tools.cu.

References SqVecSumReduction().

float getLJCutPotentialEnergy	(	FLOAT_ARRAY_TYPE *	dPos,
		float *	dEpot,
		float *	sum,
		int &	N,
		float &	boxSize,
		float &	rCutSq,
		dim3 &	dimGrid,
		dim3 &	dimBlock,
		int &	maxThreads,
		int &	maxBlocks
	)

Calculates the total Lennard-Jones energy of the system: $E_{LJ}=4\epsilon \sum_{r_{i,j}}\left[\left[\frac{\sigma}{r_{ij}}\right]^{12}-\left[\frac{\sigma}{r_{ij}}\right]^{6}\right]$ , where due to Lennard-Jones units $\sigma=\epsilon=1$

Parameters:

`[in]`	dPos	Particle position array
`[in]`	dEpot	Array on the device to hold the total potential energy for each particle
`[in]`	sum	Summation variable on the device used for intermediate data during reduction, content will be overwritten
`[in]`	N	Number of particle
`[in]`	boxSize	Edge length of the cubic simulation box
`[in]`	rCutSq	Squared Lennard-Jones cutoff radius
`[in]`	dimGrid	Grid dimension
`[in]`	dimBlock	Block dimension
`[in]`	maxThreads	Maximum number of allowed threads
`[in]`	maxBlocks	Maximum number of allowed blocks

Returns:: Returns Total Lennard-Jones energy of the system.

Definition at line 303 of file gpu_tools.cu.

References checkCudaError(), and sumReduction().

__global__ void gpuGetHarmonicPotentialEnergyKernel	(	float *	dEpot,
		FLOAT_ARRAY_TYPE *	dPos,
		int *	cnn,
		size_t	pitch,
		int	N,
		int	numNN,
		float	k,
		float	boxSize
	)

Kernel function called by getHarmonicPotentialEnergy to calculate the harmonic potential.

Parameters:

`[out]`	dEpot	Array on the device to hold the total potential energy for each particle
`[in]`	dPos	Particle position array
`[in]`	cnn	Array holding the indices of harmonic linked particles
`[in]`	pitch	Memory alignement pitch on the device
`[in]`	N	Number of particle
`[in]`	numNN	maximum number of connected neighbors
`[in]`	k	Spring constant
`[in]`	boxSize	Edge length of the cubic simulation box

Returns:: Returns CUDA error information.

Definition at line 280 of file gpu_tools.cu.

References FLOAT_ARRAY_TYPE, and gpuDistPBC().

__global__ void gpuLJCutPotentialEnergyKernel	(	FLOAT_ARRAY_TYPE *	dPos,
		float *	dEpot,
		int	N,
		float	boxSize,
		float	rCutSq
	)

Calculates the total Lennard-Jones energy of the system like getLJCutPotentialEnergy, but sums only contributions with $r_{i,j} < r_{cut}$ .

Parameters:

`[in]`	dPos	Particle position array
`[out]`	dEpot	Array on the device to hold the total potential energy for each particle
`[in]`	N	Number of particle
`[in]`	boxSize	Edge length of the cubic simulation box
`[in]`	rCutSq	Squared Lennard-Jones cutoff radius

Returns:: Returns CUDA error information.

Definition at line 314 of file gpu_tools.cu.

References FLOAT_ARRAY_TYPE, and gpuDistPBC().

Basic Observable Measurement Functions [Tools]

Functions

Detailed Description

Function Documentation

Basic Observable Measurement Functions
[Tools]