This paper presents a method for clustering scientific publications using two software tools, CitNetExplorer and VOSviewer, which are used to analyze the resulting clustering solutions. The clustering is based on direct citation relations. CitNetExplorer is used to cluster publications and analyze the results at the level of individual publications, while VOSviewer is used to analyze the results at an aggregate level. Both tools use visualizations to support the analysis. CitNetExplorer employs a clustering technique that maximizes a quality function based on direct citation relations, which is a variant of the modularity function. This technique is compared to other clustering methods and is shown to avoid the resolution limit problem. The paper demonstrates the use of these tools on a large dataset of astronomy and astrophysics publications, showing how they can be used to create and analyze clustering solutions. The results show that the clusters are relatively independent, with most citation relations occurring within clusters. VOSviewer is used to analyze the clusters at an aggregate level, showing the relationships between clusters and the topics covered by each. The paper concludes that these tools are useful for bibliometricians to perform sophisticated cluster analyses without requiring deep knowledge of clustering techniques or advanced computer skills.This paper presents a method for clustering scientific publications using two software tools, CitNetExplorer and VOSviewer, which are used to analyze the resulting clustering solutions. The clustering is based on direct citation relations. CitNetExplorer is used to cluster publications and analyze the results at the level of individual publications, while VOSviewer is used to analyze the results at an aggregate level. Both tools use visualizations to support the analysis. CitNetExplorer employs a clustering technique that maximizes a quality function based on direct citation relations, which is a variant of the modularity function. This technique is compared to other clustering methods and is shown to avoid the resolution limit problem. The paper demonstrates the use of these tools on a large dataset of astronomy and astrophysics publications, showing how they can be used to create and analyze clustering solutions. The results show that the clusters are relatively independent, with most citation relations occurring within clusters. VOSviewer is used to analyze the clusters at an aggregate level, showing the relationships between clusters and the topics covered by each. The paper concludes that these tools are useful for bibliometricians to perform sophisticated cluster analyses without requiring deep knowledge of clustering techniques or advanced computer skills.